Salesforce Admins Podcast
The Salesforce Admins podcast features real-life Salesforce Admins, product managers, and community leaders who transform businesses, careers, and community with clicks, not code. This 20min (sometimes a bit more) weekly podcast hosted by Mike Gerholdt feature episodes to empower Salesforce Admins who are implementing Enterprise CRM solutions. There may be some (digital) confetti. For more than our most recent episodes, go to https://admin.salesforce.com/salesforce-admin-podcast.
info_outline
How AI Could Change the Way Admins Set Up Salesforce
08/20/2026
How AI Could Change the Way Admins Set Up Salesforce
Today on the Salesforce Admins Podcast, we talk to Khushwant Singh, SVP of Product Management for the AI Application Development Platform at Salesforce. Join us as we chat about how rethinking the Salesforce stack for AI with Headless 360 will change the way admins set up complex systems. You should for the full episode, but here are a few takeaways from our conversation with Khushwant Singh. Opening Salesforce to AI agents AI is changing more than how we find information or generate content. It could also change how Salesforce Admins configure the platform itself. I sat down with Khushwant “Khush” Singh, to learn about Headless 360 and the work happening behind the scenes to make Salesforce accessible to both humans and AI agents. “Headless 360 is about bringing Salesforce to wherever you need it,” Khush says. In a headless platform, he explains, the frontend is separated from the backend. So you can use Salesforce data and visualizations on your website without changing the design, or change records in Slack without having to open an extra tab. But Salesforce Headless 360 is about more than portability. Khush and his team are translating the entire platform into a set of model context protocols (MCPs), skills, and metadata that can interface directly with AI agents. And that opens a world of possibilities for what admins can do with conversational tools. What if setting up Salesforce became a conversation? “For the last 27 years, layers that make up the Salesforce platform have been put in place for a human,” Khush says, “so we’re trying to open this up for both a human and an agent, so the agent can do that work on behalf of the admin.” Setup with Agentforce, which streamlines configuration for admins with the help of an AI assistant, is a great example of what Khush is talking about. You can use natural language prompts to create objects, Lightning pages, and flows; manage users and user access; and troubleshoot formulas. Headless 360 lays the groundwork for a future where setting up Salesforce becomes a conversation about your business goals and processes. No coding required. Admins still stay in control None of this removes the admin from the process. Khush is quick to point out that the goal is for agents and admins to work together, with humans still initiating actions, validating results, and providing governance. His advice for admins right now is simple: try Setup with Agentforce and start getting familiar with what this new way of working feels like. Because the biggest shift may not be AI helping you use Salesforce faster—it may be AI changing how you build Salesforce in the first place. Listen to the full episode for more from Khush about Salesforce Headless 360 and how AI could change the way we set up our orgs. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce 360 Blog Post: Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Salesforce Help Article: Salesforce Admins Blog Post: Salesforce Admins Blog Post: Trailhead: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: This week on the Salesforce Admins Podcast, I'm joined by Khushwant Singh to talk about what happens when Salesforce stops just being a destination and starts becoming part of every system your business already relies on. That's right, from Headless 360 to Setup with Agentforce, Khush breaks down how admins are moving beyond configuring features and into orchestrating how data, AI agents, automation and people work together across the business. Now, we're going to dig into what Headless actually means for admins, why governance and trust matter more than ever in an AI first world, and how Salesforce is building toward a future where setting up complex systems really becomes more of a conversation than a checklist. This episode is really about systems thinking, how admins can connect experiences without forcing teams to rebuild everything from scratch. So if you've been trying to figure out where AI fits into your role or how your role is evolving alongside it, I promise you this episode gives you a practical look at what's already happening and what's coming next. So be sure to hit that subscribe button. Share this episode with your favorite architect or admin friend, and let's get Khush on the podcast. So Khush, welcome back to the podcast. Khushwant Singh: Thank you for having me back, Mike. Mike Gerholdt: I say that because I had to look and it was 2022 back in the day you were talking Experience Cloud when we last had you on. You've since moved on, moved up, had your hands in a lot of things, did an admin keynote with me. So let's start there and let's kind of catch everybody up on some of the products you've been working on, and then we're going to talk about the newest one that we launched at TDX this year. Khushwant Singh: Yeah, happy to. And again, great to be back. I think back in the day, started obviously with Experience Cloud. Experience Cloud really was building a product that is external-facing and again, it's built on the same platform. And again, as we looked at the product portfolio, it just made sense to start to bring things together. So we've actually sort of rolled and morphed into the application development platform, obviously within the platform organization, everything pro-code, low-code, desktop, mobile. And so there've been quite a few products leading up to here, anything from our delivery on mobile, mobile offline, the Salesforce mobile app, all the way to Agentforce Vibes, the Agentforce Experience layer, obviously a fair amount of work on the Headless side of things. So yeah, it's been quite a journey to this particular point, Mike. Mike Gerholdt: Yeah. I mean, it's actually hard to point out things that aren't admin-friendly that you've had your hands on because you've always kind of championed everybody being able to build on our platform. So let's talk about TDX this year because we launched Headless 360. Khushwant Singh: Oh yeah, we did. Mike Gerholdt: By the way, in case you forgot. Khushwant Singh: Yeah, it was a massive announcement. I think, look, as you think about what's happening in the market right now, you've got customers engaging through many different surface areas. We've got customers on ChatGPT, on Claude, on Gemini, Teams. At the same time, you've got developers also using a variety of different tools, different IDEs, different coding agents. And so I think Headless 360 really is a testament to acknowledging this change in the market in the industry and trying to meet the customer where they are. The customer being an end user or the customer being a developer and admin, just meeting them where they are in whichever tool, whichever interface they might be. Mike Gerholdt: Yeah. I mean, do you recall ever a time in technology when it was, I hate to use the euphemism, but like the Wild West where I feel like there's so many choices for what you can do just in regards to AI? Khushwant Singh: There is. And there is obviously a good side to it. There's also a downside to it. The good side is really your [inaudible 00:04:31] for choice. There is innovation that is helping you drive productivity day in, day out. The downside of this is you just have to be really careful about quality, about trust, about governance. And you have to ensure that while you're going fast, you're not sort of compromising on that trust aspect of things. And that's what we're trying to do here at Salesforce and just trying to be a little bit more thoughtful in our approach over here. Mike Gerholdt: Yeah. So my developer friend, Josh Burke, who comes on the podcast, usually has to explain things to me every now and then. I think developers are very familiar with the term headless. Khushwant Singh: Yes. Mike Gerholdt: I'm not sure outside of Headless flows, which I know what those mean, I'm not sure admins could understand or could confidently buzz in on a game show and answer what is Headless 360. So can you help them kind of understand what is Headless 360 if they had to tell their manager? Khushwant Singh: Yeah, sure. So one way to think about it is, let's say you're a customer and you've got your own website, it's built out of React, it's not built on the Salesforce platform, but there is so much of your goodness and investment that you have within the Salesforce platform that you want to be able to easily reuse in the context of this other website that you have, portal that you have, mobile application that you have. All of that may not be built on the Salesforce platform. You want to be able to reuse your investment seamlessly. And so Headless 360 really is about how do we empower you, the customer, to be able to easily use those assets, those investments within the Salesforce platform in that surface area that you've got that may not be on the Salesforce platform. So it's really about giving them that flexibility so that they can easily reuse their investments in all of these various surface areas and just meet their customers wherever they are versus bring their customers into Salesforce. Mike Gerholdt: Yeah, I think it makes sense. I mean, I was a customer way back in '06 when I remember it was like transfer and move to Salesforce, and everything was "have it on the platform, have it on the platform." And then shortly thereafter it was kind of like almost the realization that we had was, "Well, it's a lot of friction to constantly move platforms. What if companies just integrate and we become that connector to everything?" And I feel the same way with this. Khushwant Singh: 100%. And you know what? Look, I think the one example is if you go into the Williams Sonoma website, right? Williams Sonoma website, it's not built in Salesforce, but if you use their sous chef AI, I think... I'm not sure what they call it, but- Mike Gerholdt: I'm sure it's something really, really snappy and small. Khushwant Singh: Olive. Olive. There you go. Mike Gerholdt: Olive. Yeah. See? Khushwant Singh: A snappy name right there. That agent is an Agentforce agent, right? So that's on example where Salesforce is being used in whichever head that you'd like us to be in. And so in this particular case, the Salesforce agent is in the Williams Sonoma website. So that's just one example where we are trying to make it easier for this integration to happen through the Headless 360 platform. Now that's one view. And that's the view, I would say, an easy way to try to understand what it means for Headless 360. Mike Gerholdt: Yeah, because then as an admin, as a developer, you're not having to sit down and say, "Okay, well, if we move to the Salesforce platform, we have to deprecate everything that we just spent the last two years building on our site." Marketing's like, "No, no, we just got the site pixel perfect the way we want it. Now we can integrate best of everything that we want." For admins who aren't thinking of maybe websites, how does Headless 360 play into other integrations that they may talk to IT or internal stakeholders about? Khushwant Singh: Yeah. So for example, when admins are chatting with their other internal team, they may have a custom application that they already have in place and they want to be able to have Salesforce data, Salesforce visualizations surface in whichever application they may have already, right? That's one example. So again, it is about bringing Salesforce to wherever they may already have. So for example, an admin may have a business application that already runs within ChatGPT. So ChatGPT, for example, it allows companies to create bespoke applications. And if you want to be able to integrate Salesforce into that application, that's doable as well. It doesn't have to be in ChatGPT. It could be in any other existing non-agentic application as well. But it's about that, look, you don't have to rebuild everything. Wherever you are, you've got an existing application, we'll find a way through Headless 360 to be able to bring Salesforce to it. Mike Gerholdt: Yeah. No, it totally makes sense. I mean, in that way we can really think about how are we making sure that the departments have the software and the tech that they want, but also access to all of the information that they need to do their jobs the best. Khushwant Singh: Mm-hmm. Mike Gerholdt: So Headless 360 is a big thing we talked about. I also know I saw you in some True to the Core and True to the Core Deep Dive stuff. Was there some features you were talking about in that that possibly play into what admins are going to be working on for the rest of this year? Khushwant Singh: Well, yeah. I mean, a few aspects as we think about Headless 360 that would be of interest to admins. So first and foremost, you've all seen the stack diagram of the Salesforce platform. You've got those four layers. You've got your layers of engagement, you have layers of context, you've got the sort of our data foundation, et cetera. So you've got those layers that make up the Salesforce platform. For the last 27 years, those layers have been put in place for a human, right? And now what the product teams and the engineering teams are doing are going back and saying, "How do we now open this up both for a human and an agent so the agent can do that work on behalf of the admin when the admin sort of wants a task to be done for that particular reason?" And so what that means is that we're translating our entire platform into a set of MCPs, a set of skills, a set of metadata that is grounded so that these agents can do the work on behalf of the admins. So one good example is the setup with Agentforce, which is so far so good. We've gotten great feedback on it. Now behind the scenes is a set of skills, a set of MCPs that are firing off to be able to get their tasks done. And so that's one thing that I think our admins can look to benefit from. The other piece is I know we have some admins who are familiar with some of these IDEs that we have out there. They use Cloud Code quite a bit. They use Vibes quite a bit. It may not be the entire admin audience, but for those audiences, again, the ability to be able to use all of these MCP skills in these development environments is another thing that our admins can really look forward towards. Now the other thing that we've also worked on is, over the last few months, the foundation for these skills, these MCPs, this Headless 360 foundation has been very sort of focused on how do we unblock these pro-code use cases, whether that's in Cloud Code, whether that's in Vibes, et cetera. But this is the same foundation upon which as we speak in safe harbor, I share that we are building what we call our no code offering as well. So how do you go about in an unintimidating sort of type of interface, have this conversation with an agent, which behind the scenes is interacting with that same Headless 360 layer to help you build out your agents, your applications, to help you augment your existing agents and applications, help you set up provision, configure Salesforce from a zero to one point of view, because we all know how setting it all up is quite a taxing task for our admins? So that's the potential I see for Headless 360 for our admin personas. Mike Gerholdt: Yeah. I mean, to go back on some of the stuff that you were mentioning, I remember the first time I thought, "Hey, I could have ChatGPT rewrite ChatGPT's instructions." And I remember thinking like, "I'm so far ahead of the curve, man. I'm making AI rebuild AI." And then it was only a few weeks later that Cheryl was demoing Setup with Agentforce and I thought, "Oh, that makes so much more sense because all of the things that you were having that setup agent do are things that not necessarily admins don't like doing. It's like paper cuts." It's the little things that take 20, 30 minutes out of your day when you really wanted that solid hour to build that application kind of start to finish. Khushwant Singh: Indeed. Indeed. So we really viewed Headless 360 as the foundation upon which agents can work on behalf of the admins. And again, look, this is agents and admins working together because again, we've got to have the human presence to be able to validate, to be able to interrupt, to be able to initiate any actions that are performed by the agents. And if these agents can help drive greater productivity, then it's a win-win while at the same time having governance by humans over it. Mike Gerholdt: Right. Absolutely. Do you envision us being at a point where all of setting up Salesforce could just be a conversation with an agent based on the business and what they're trying to accomplish? Khushwant Singh: We are trying to do that. We are trying to do that. Now, I'll give you a view into our approach over here. We're taking a steel thread approach here where just trying to, for example, set up a customer service agent. That cuts across Service Cloud, cuts across platform, cuts across Agentforce, cuts across Data Cloud. So we've got to take a system view, an end-to-end view versus a feature view because that's what our admins go through. That's what a practitioner goes through. They look at it from a steel thread point of view, and that's what we as product teams have to do. And so as we think about using agents to go ahead and help provision, set up, get Salesforce up and running really quickly, that's the new strategy that we are taking over here, a steel thread view so that we can look across the board and say, "Look, which MCPs are missing? Which skills are missing? Which APIs are missing? Which metadata needs to be grounded? How do we ensure that across entire steel thread, quality is being ensured, governance is being woven into?" So we are starting that way and we want to scale that way as well. So my hope is that come Dreamforce this year, we'll be able to at least have a sizable number of steel threads that we feel really, really confident about on Agentforce really helping and agents helping in this particular side of things. So yeah, that's the approach we're taking, Mike. Mike Gerholdt: Wow, that's going to be amazing. Every time you think, "Oh, we've probably invented everything," technology just comes along and pulls the rug out from underneath of you and says, "Haha! There's new stuff. You got to figure this out now." I'd be curious, you're so ingrained in a lot of things that admins do on the platform, a lot of things developers do on the platform and then also having to keep up with AI. What is a piece of advice you could give people for how you personally keep up with the newest news and innovation on what's going on with AI? Khushwant Singh: Wow, that's a good question. I think first and foremost, I spend a lot of time just looking at reading publications like what's on TechCrunch, what's on Techmeme, trying to just understand how are technology providers adapting, how are companies using AI. That's one. The second one is actually you learn a lot via speaking with others in the community, right? Mike Gerholdt: Mm-hmm. Khushwant Singh: You talk to customers, you talk to other admins, you talk to other developers and you get to understand there are use cases that they're working through and some of the very many innovative workarounds or users of AI that they're applying. And then you're thinking about this that, "Wow. There's a trend here, there's something that we should be backing up and maybe providing out of the box within the product." And so I think that community piece speaking with other customers, with your peers, your colleagues from...
/episode/index/show/buttonclickadmin2/id/42490025
info_outline
Stop AI Hallucinations with Better Salesforce Metadata
08/13/2026
Stop AI Hallucinations with Better Salesforce Metadata
Today on the Salesforce Admins Podcast, we talk to Skye Tyler, Principal Solution Consultant at Beyond the Horizon Technology. Join us as we chat about why clean metadata is crucial for getting the most out of AI. You should for the full episode, but here are a few takeaways from our conversation with Skye Tyler. Why metadata is your training manual for AI You’ve done it, I’ve done it. Sales wants a name changed on a report, so you just change the field label rather than create something new and mess with the integrations. Or maybe you have four different dates, but your marketing team knows which one they’re actually looking for. What’s the big deal? The problem is that while humans can get by on institutional knowledge, AI agents need a bit more help from your metadata. As my guest this week, Skye Tyler, explains, you should think of AI like an overenthusiastic intern with short-term memory problems. It really wants to answer your question, so if it sees four different dates, it’s going to take a guess. Think of metadata like an onboarding manual for AI. With clear field descriptions and help texts, you can point your agents in the right direction and prevent hallucinations. How to get started with metadata cleanup Obviously, cleaning up your org’s metadata is easier said than done. However, Skye has some great tips for how to get started. If your business has busy periods and slow periods, you might be able to schedule time specifically to work on your metadata. But for the rest of us, Skye recommends a clean-as-you-go approach. Any time you update an object, set aside some time to review all of the related metadata and update the description. Similarly, any time your organization is building a new business process that includes AI, you need to clean up any metadata an agent could interact with. As Skye points out, while you don’t necessarily have time to deep clean your house every week, a little spot cleaning can go a long way. Skye’s Salesforce journey and the value of sharing what you know We also took some time to talk about Skye’s career path from nonprofit arts and what advice she has for folks who are new to the Salesforce ecosystem. AI is changing things fast, and that means there are more opportunities. “With all of these newer technologies and tools, anybody who's looking to come into the space today is learning the same things that I'm learning, having been in the space for a decade,” she says. The Salesforce community is incredibly supportive, and Skye recommends putting yourself out there. You never know who you’ll meet at an event, and where those connections could take you. “No matter where you are in your journey, you know something that other people don't,” she says, “put yourself out there, and I guarantee you someone's going to appreciate it.” Listen to my full conversation with Skye for more on how to clean up your metadata and stop AI hallucinations. And make sure you’re subscribed to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Salesforce Admins Blog Post: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins Podcast, I'm talking with Skye Tyler about the part of AI readiness that, well, it's easy to overlook. That's your metadata. We all know that clean data matters, but agents also need clear field descriptions, documentation, permissions, and context to understand what your business actually means. Skye explains why metadata acts like onboarding for an enthusiastic new employee and why a confident answer isn't always a correct one. We'll also talk about practical ways admins can improve metadata without trying to clean the entire org overnight. So give this episode a listen, subscribe wherever you get your podcasts, and of course share it with an admin who maybe has prepared their platform for Agentforce. And with that, let's get Skye on the podcast. So Skye, welcome to the podcast. Skye Tyler: Hi, thanks for having me back. Mike: I'm excited. Going through everything that we're doing to prepare for Dreamforce and admins are preparing for the agentic future, which is coming faster than eve, I stumbled across your session. And I thought, hey, before we get into Dreamforce, maybe we can dive into all of the stuff that you might not be able to cover in your 40-minute breakout session. So we'll talk about that because it's data related and data related is agent related. But before we get started, can you just refresh everybody on how you got started with Salesforce and what you do in the ecosystem? Skye Tyler: Oh, I'd love to. So I've been fortunate enough to be working in the Salesforce space with the nonprofit industry for a little over a decade now, and started off doing implementation work with nonprofit arts organizations, doing ticketing and donation. And then I've moved in and through the consulting space. Again, focusing primarily on nonprofits with a little bit of dipping into education along the way. And over the course of that decade, I've been fortunate enough to meet some amazing folks, yourself included, that has given me an opportunity to have the chance to share not just my journey, but my experiences, my struggles, successes, failures. And some of those avenues have included things like coaching and mentoring with super moms and the Salesforce military program, and then also just being really honored to be able to speak at community dream and events and events like Dreamforce and Trailblazer DX. Mike: I mean, that's a great answer. You know what's funny is so many people ask me, "How do I get started in the community," or, "What should I do?" And I feel like now I'm just going to point back to your answer because what you said was just perfect. Well, don't be afraid to share what you've learned, or what you're learning, or how you've learned it, or things you're working on because that's what people love to see. They love to see how you're doing it so that they can get an idea for how to do it. So thanks for answering a question I didn't have. Skye Tyler: Yeah, absolutely. I also just want to throw in, and I say this at many of my live sessions, that no matter where you are in your journey, you know something that other people don't. And everybody learns more when folks show up, and are vulnerable, and are willing to share those things. So definitely just give it a try. Put yourself out there and I guarantee you someone's going to appreciate it. Mike: Yeah, absolutely. Plus the thing I add to that is there's probably somebody out there like you that really needs to see somebody like you on stage. Skye Tyler: Yes, absolutely. Mike: So we talk a lot about data with agents and cleaning data, making sure your data's up-to-date, making sure your data's fresh, which every time I use the word fresh, I always think of how that plastic comes off an iPhone. I don't know why, but that's where my brain goes, because agents and AI and LLMs do a really good job of consuming a vast amount of information and giving it back to you in what you asked for. I think one of the things that we talk less about and that I want to talk to you about is your metadata, because I'll be honest, I pay attention to my metadata for five seconds when I'm creating the field. Skye Tyler: I'd say that's four seconds more than most people. Mike: Right, because all you got to do is click outside the box and it fills in the developer name, so I'm good. And description, ah, fill it in later. Skye Tyler: Yeah. And for the longest time, that's been really standard. And again, working across several consultancies and with lots of different works, people assume, well, we know what this field means. You give it a field name, the team knows what this means. I don't need to fill in description. I don't need to fill in help text. It's fine. And again, for the longest time, that institutional knowledge was enough of a stopgap that it was fine. Nothing was going to break or go horribly wrong if you left the description field empty. But all of that changes when you start introducing agents. And an example I may reference a couple times on this podcast is, I don't know if you've ever heard this, but the idea of thinking about your agent as a really enthusiastic intern with short-term memory problems. Mike: Like memory of a goldfish. Skye Tyler: Yes, exactly. So if you've got a new intern and they don't have deep industry experience, or they don't have longer term institutional memory within your organization, and they come on site, they're excited to be there, they're energized, they're probably young and enthusiastic, and they're going to sound really confident. You're going to ask them a question. They're like, "Yeah, absolutely. I can absolutely do that." But there's a difference between sounding confident and actually knowing the answer. And where AI gets really tricky is not only does it sound confident, it usually has a certain level of fluency to it with the amazing jumps in natural language processing over the past few years, it sounds both confident and right because it's just a well-structured sentence. And there's a huge gap between the structure of the sentence and the actual understanding of the data behind it. Mike: Right. It reads to us as so confident in its own answer that we would be silly to think it's wrong. Skye Tyler: Oh, absolutely. It would be a waste of time to go back and check it because look at how confident it is. Mike: Yeah. And look at all the part that I always think of is look at all the information it gave me. I asked it two sentences and it gave me four paragraphs of an answer. It must be right. Skye Tyler: Yes. And of course, don't forget that at the beginning, it also flattered you and said, "Oh, Mike, that's a great question. Let's dig into that." Mike: Yeah. You sound just like my AI. Skye Tyler: Exactly. Mike: "Of course it's a great idea, Mike. You should totally do that. Everything's a great idea." So when we think about that though, why is it making those mistakes? Because if we're paying attention, and I mean, I was even working on some content today that's like, got to keep your stuff fresh. If you're going to do call notes or have a summary agent, then make sure your salespeople are doing call notes and keeping it up to date or keeping the data up to date. Then why is it so confidently giving us, I don't want to say the wrong answer, but confidently providing us with information that it thinks is correct, but isn't? Skye Tyler: Yeah. So a big part of this comes down to the difference between how humans think and form answers and how the large language models do it. And one of the real risks around this false fluency framework is that you pose a question, you ask a question, and first there's the basic understanding that the agent, whichever model or system you're using, their goal is to accomplish the thing that you asked. So the first thing is getting really careful about how you ask your questions so that you're not sort of setting it up to just affirm the thing that you have asked. But then beyond that, the way that these LLMs, these large language models work, is it's basically a really complex pattern completing process. And so you mentioned peeling that plastic off of your new iPhone. How many times are you typing a text message on your iPhone and it has that auto recommend that comes up right above your keyboard? Mike: Yeah. Skye Tyler: So that's a good example of a simple version of this. So depending on how often you type in "Oh my," and then it's going to recommend a couple of different things that often come after "Oh my" in the training data. When we're thinking about this in the context of asking an agent something in Salesforce, the way that it identifies what to propose for an answer is not looking at just the data in the field, it first has to go look and see what information is available and that's when it dips into the metadata. So if you've got four different fields that have very similar field names, but have no metadata to tell the agent field one means this, field two means that, field three means this, and field four means this completely other thing, it's just going to guess. Mike: So if you had four dates on a record that were, I don't know, draft dates because it ties to a contract or something, and it's draft one, draft two, draft three, draft four, and you know as the user, those are the dates that you completed those drafts, but your metadata just shows four draft dates and there's no explanation for the agent. So the agent's like, "Here's when all the drafts were completed." Skye Tyler: Right, it's going to guess. And because it's non-deterministic, it might guess right the first time you ask it a question. It might guess right the second, third, fourth, fifth, 20th time, and then the 21st time, it guesses wrong. Mike: But confidently. Skye Tyler: But confidently wrong. Mike: Confidently. Right. Skye Tyler: Right. And so when we're thinking about how to support our agents, if we go back to that analogy of the over-enthusiastic summer intern, the metadata is effectively the onboarding process for that new intern or a new employee. It's the manual about what your data means, and it's the context that the agent can use to give you more consistently reliable, correct answers. Mike: Love it. So I remember, I think it was TDX two years ago, we were talking kind of on this subject. And whenever we roll out a new feature, it's here's the things you need to do to prepare. And I remember getting up and talking with a lot of people of like, "Hey, it's more than just the data that's got to be cleaned because there's a lot of data cleaning services and hopefully you have a lot of validation rules and stuff in place, but also your metadata needs to be cleaned as well." And I looked over in the audience, I felt this guy just shrug his shoulders and just kind of turn a lighter shade of pale, I'll say. And I walked up to him afterward and I asked him, I was like, "Are you okay?" And he's like, "Yeah, your words just hit me really hard. I was thinking of all the metadata I have to clean up." So obviously it would be awesome if every day people woke up and you cleaned your whole house. But let's be honest, you don't have to wake up every day just to live in your house to clean your whole house. What is the approach that an admin should think of in terms of getting their metadata to that next level and to prepare for an agent force implementation? Skye Tyler: Yeah, I love the analogy of cleaning your house every day because I'll tell you, I do not wipe my baseboards every day. Mike: oh, man. I would pay somebody to come once a week to do my baseboards. It would be awesome. Skye Tyler: Yeah. So there's a couple of ways that you can look at this. The first one is figuring out within your workflows, your day-to-day and annual workflows, where you can build metadata maintenance into your process. And so what this might look like for admins is the next time someone on your team asks for an update on an object, whether it's a page layout, or some new fields or what have you, taking and building some extra time into however long you budget or allocate for that task to include reviewing related metadata. Every time you need to update a flow because the business process has changed and you say save as new version, as soon as you click save as new version, update the description of the flow. That's the first thing that agents are looking at to be able to figure out what automations do I call, what automations are related to a process, et cetera. So building it into the workflow of what you already do is one avenue to consider. If you have an amazing organization that lets you do proactive maintenance, working that into your project plan and saying, "Hey, this is a usually kind of quiet time of the year for new admin requests. Let's go ahead and block two weeks or three weeks and do a more thorough or a deeper dive into updating the metadata." If you work at one of those organizations, let me know, because I've heard they exist. I haven't encountered one in the wild. But then lastly, when you have any new agent related process, it should be considered mandatory as part of that project to do a metadata review of every aspect of metadata that that agentic process is going to interact with. Because one of the big challenges that organizations have been seeing is they're putting a lot of time and resources and money into creating agentic solutions, but the research and the data is just not showing the level of return on investment or the level of success that people are expecting. And one of several underlying factors is you're not setting your agents up for success. You've got this new employee, this new agent employee, but you didn't onboard them. You didn't provide them with adequate institutional knowledge and training. You gave them some direction. But if you think about it, Mike, when you've got a new employee that comes on and they go to that podcast or that blog record for the first time and they see your four review dates and they don't know, they stop and ask someone. They ping you on Slack and be like, "Mike, can you explain these four review dates to me?" Your agent can't do that. So set them up for success with that onboarding through the metadata updates. Mike: And I think that also, I mean, I really like how you approach that. I think that's also a funnel or a pathway that you could flip. So you don't have to think of, I have to tackle everything and clean all my metadata first, as opposed to, well, if we're going to roll out an agent and it's going to touch these three objects and these 15 fields, then that's where you start. Skye Tyler: Absolutely. Mike: Because then that will help you set up a plan for here's how we're going to tackle this part of it. And then as the agent responsibilities or you build additional sub-agents expands, then you can tackle two or three more objects that maybe that sub-agent has to talk to as opposed to, okay, we're going to wake up at 3:00 AM, clean the whole house just to live in it, which is a good idea. I think sometimes readiness to the extreme can feel overwhelming. Skye Tyler: Right. And if it's overwhelming to the point where you don't make any progress on it, then it's not useful either. So definitely thinking about what's within scope. Where I would challenge that a little bit though is unless you have put very firm guardrails on your agent, there's always the possibility that it's going to get creative. So you might intend for that agent to only look at these four fields, or these 15 fields, or these two objects, but if you don't actually instantiate that guardrail or that limit into how you build the agent, it could start to be like, huh, I wonder if this related record might impact me being able to answer Mike's question better. Let me just go take a look at that. And that's where we could get into some of the more deeper technical things, but the idea that authentication is not the same as permission. And so making sure that each of your agents has very clearly defined, and that ties in with the metadata too, because all of those settings in Agentforce and Agentforce builder are set up, the sub-agent descriptions, the scope, all of those things are metadata around your agents. Mike: And I think there's downstream effects to this that are also positive too. I'm thinking of all of the reports that you're dealing with that will suddenly get better because I know you're always building reports, sometimes not off of the report label. So help me through this. You've got an implementation and you have an integration with another system and you change metadata descriptions or field names. How do you need to...
/episode/index/show/buttonclickadmin2/id/42412745
info_outline
From Inbox Requests to a User Management System
08/06/2026
From Inbox Requests to a User Management System
Today on the Salesforce Admins Podcast, we talk to Michelle Wolfe, Platform Engineer. Join us as we chat about how she built a Flow-powered user management system to handle requests, approvals, user creation, permissions, communications, and even onboarding. You should for the full episode, but here are a few takeaways from our conversation with Michelle Wolfe. Streamlining change requests with screen flows It’s a good thing when your organization is growing, right? For Salesforce Admins, however, new hires need new accounts, and that means approvals, permissions, and onboarding. My guest this week, Michelle Wolfe, found herself in exactly this situation. She turned to Flow to build an automated user management system with no code, and she’s here to tell us all about it. Michelle’s company was bringing in 80+ people per new hire class, but the account creation request process was a mess. Managers would email a random member of the three-person admin team, and then they would manually create the account. So the first step was to wrangle the business process with a screen flow to get the correct information into Salesforce and create a case. But Michelle was just getting started. Now that everything was in Salesforce, she knew that she could use Flow to automate the rest of the process. New account creation with autolaunched flows Once the new user’s information was approved by their manager, Michelle set up an autolaunched flow to spin up the account. This took care of data validation, configuring the correct permissions, and creating a username. Finally, a second autolaunched flow would close out the case and send two email actions: one to confirm the request, and one to welcome the new user. With help from Flow and Einstein Copilot, Michelle transformed a complicated manual business process into something streamlined, simple, and scalable. Partner with your training team for a fresh perspective on your org Michelle credits her success with how closely she works with her training team on enhancements. “Because they interact with new employees, they see things differently than someone who’s been here for years and knows how we use our Salesforce,” she explains. That fresh perspective helps her identify improvements she might otherwise miss. Listen to my full conversation with Michelle for more on how she used Flow to automate user management—we really get into the weeds. And make sure you’re subscribed to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins Podcast, I'm joined by Michelle Wolf to talk about what happens when user onboarding stops being a string of emails and becomes a real platform process. So Michelle built a Flow-powered system that handles requests, approvals, user creation, permissions, communications, and even offboarding, while keeping the admin team in control. We're going to talk about the automation behind it, but also the process decisions that make it accurate, secure, and useful for her business. Now, because when admins design access and onboarding well, they're not just saving time. They're protecting trust and helping people get productive faster. So listen in, subscribe, share this episode with an admin who has one too many user requests sitting in their inbox. I'll tell you who doesn't have too many user in requests sitting in their inbox, and that's Michelle. So let's get Michelle on the podcast. So Michelle, welcome to the podcast. Michelle Wolf: Thank you for having me, Mike. Mike: I think this is exciting because I can't recall the number of times I've had people on about onboarding users. And I think it's because whatever cool tool we have come out, it's the second thing people do. They're like, "Oh, I could do this, and I need to figure out how to do X, Y, and Z when I onboard people." Because it was like that with Flow. There was stuff with Chatter when that came out, "How do I onboard people and add them to groups?" And Jennifer Lee was at your session at Midwest Dreaming, and she thought it was just phenomenal. So I had to have you on the podcast to talk about it. But before we get into that, tell me a little bit about yourself, how you got started with Salesforce and what you do. Michelle Wolf: Yeah. So my journey started as a lot of ours as an accidental admin. I was working for a small family-run business who had Salesforce, and they were paying a consultant to basically be their full-time admin. And one day my manager walked into my office and said, "I don't want to pay them anymore. I need you to learn this." And I was like, "Okay." Mike: "I suppose." Michelle Wolf: Yeah. Why not? So yeah, that's what started my journey. And I skilled up on Trailhead and I was on it two, three hours a day trying to figure out how to be an admin in Salesforce. And this was back at the initial transition from Classic to Lightning. Mike: Ooh, fun. Michelle Wolf: So half of my early batches are on Classic. There just wasn't a ton of Lightning stuff out there yet. So I would build a lot in Classic and then flip it over to Lightning to see what it looked like because they did like the Lightning interface, which was great. So that started my journey. And then when I decided to part ways with that company, I took a different position and paused my admin career. And then about five years ago or so, I had an opportunity to be a product SME for Salesforce. And I was like, "Yes, I miss this product. I miss doing this. I miss being able to make things better with just the click of the button. So let's go do this." And I jumped in both feet and never looked back. And so I've been an active admin for a little over five years now. And I'm a Flownatic. Anything I can automate, I am doing it. Mike: I love it. Active admin as opposed to passive. Michelle Wolf: Exactly. Mike: I don't know. So I love when you started because I remember those times and the reason there wasn't enough Lightning things is we were all writing it as fast as we could. Michelle Wolf: Yeah, I know it. Mike: It's kind of like learning to drive with a manual. And then they're like, "Oh, well, here's an automatic." And you're like, "Oh, I already know how to drive with a manual." I mean, if you knew how to do things in Classic, then doing stuff in Lightning was just that much easier, in my opinion anyway. Michelle Wolf: Yeah. I agree. I agree. And the user experience was better for my users. They liked the layout and the feel. It jived with their personalities better instead of that very formal grid-looking everything. Mike: I mean, that was the internet at a certain period in time. And then it grew up and it was like your eyes don't have to squint with white space. Michelle Wolf: Right. Mike: Every field just went from tight polyester pants to sweatpants. And it was like, "Yay." Michelle Wolf: Exactly. Mike: Let's talk about being a Flownatic because I've talked with Jen. I know Jen said she was in your session. I really think, I mean, outside of just it's so incredibly powerful, the amount of things that you can do on the Salesforce platform without writing a single line of code. Flow is one of those. And I remember it was 100 years ago that I saw Flow for the first time at Dreamforce when I was a customer. And I think they called it Business Process Management or something. And it was an app you had to download, and then you had to upload schemas to it. It's since graduated into a much more robust tool. But the nice thing is once you know that, I mean, caveat emptor because AI runs off of all the flows. I mean, if you're good at Flow, there's very few things you can't make Agentforce do. So tell me a little bit about why you built onboarding with Flow. Michelle Wolf: Yeah. The company I was with found themselves in a really big period of growth, which was really amazing. But the new hire classes were like 80 people. And I just could not bring myself to want to spend a week of my life just filling in these boxes to make new users. It was really so time-consuming. And one typo here, one missed field here. My validation rules didn't check out. It was just a pain. It's just very mind-numbing. And I was like, "There has to be a better way." And I went through the different variations of the Add Multiple Users. It's an out-of-the-box function, except I couldn't put my custom fields on there, which means I couldn't save my users because we had validation rules in place. And then I experimented with a bulk upload, but it was such a pain because if I would typo a field, I would mess up my entire formula. Mike: Yeah. Michelle Wolf: And I was like, "There has to be a better way. There just has to be a better way." And- Mike: Plus then all you're doing is just creating the user. Michelle Wolf: Exactly. Mike: With Flow, you can do so much more. Michelle Wolf: Exactly. So by converting it to a Flow, I had a screen. I used my screen flows, and I leveraged the fact that the onboarding team is already putting that information into an email or something. Well, just go here and put it in a case for me. Just hit this button, pull out the fields. Now I don't have to do this. And then I used that screen flow to make a case so I can track my productivity, when the requests are coming in, when the requests are due by, who's doing the request. We started with just managing cases manually and making the users off the cases. And we had built in the approval process to make sure that someone else was double checking the inputs. Did they spell their name right? Did they put in the email right? Because I don't know these people. They're not going to be reporting to me. I'm not part of the onboarding team. I'm just making a user. So we had the managers do the approvals and make sure all that information's correct, make sure they're requesting the right... We use a field called Team Name that we made custom to make sure they're getting put on the right team, the customer service team or senior customer service team, whatever. So someone else is double checking it and then approving it. And I was like, "From here, I have to be able to automate something because the information's already in Salesforce." And that's when I got into utilizing that approval flow to trigger an autolaunched flow to actually make my user and create the alias and create the username and update the profile based off of what was entered on the request form in my screen flow, and just physically make my user. Mike: Wow. We spent a lot of time talking about the tech part of it. I want to dive in because I feel like you conquered a lot, but you probably had to sell it. Maybe you didn't. Tell me about all the non-tech stuff that you had to sell. Who did you have to go to, and what were the people you met with in order to say, "Here's how I'm going to onboard users moving forward"? Michelle Wolf: So the onboarding process had already been moved to our team. And it was just that it was coming to us in a very casual way. We're getting an email request. So it was part of the struggle was they started emailing us individually. We were a team of three, and they would just pick which admin they wanted to work with. Well, if I'm out of office or I have to leave unexpectedly and forget to put my out of office on, then their stuff's getting delayed. And so that was the first big conversation of how do we take out that portion? How do we get them to just put it in so we can decide who does the work? Because if I'm heads down on a project, I don't have five minutes to go make a user. I need to be heads down on my project. Mike: Right. Michelle Wolf: So this allowed us to divvy up work more easily and really even just check our own productivity of handling these types of requests and making sure it's not one person that always gets stuck with the work because they're the one that's always being emailed. So that was the first thing. And so the three of us, we were a really tight-knit group, and we're all very aligned with not wanting to be the only human that was pinged or emailed. And so it was like, let's just get it to a centralized place. And so we just went to our manager and said, "We're going to build this. We need this to be more productive." And they were like, "Yeah, makes sense. Do what you need to do." Mike: I mean, that's awesome. Michelle Wolf: Yeah. Mike: Sometimes when I was an admin would happen and sometimes it wouldn't. But I do think some change has to happen that way of, "No, here's how I'm going to be more productive, and here's what I want to set up." Michelle Wolf: Yeah. Mike: So then were there... I guess what I'm getting at is, for admins that hear this and like, "Oh man, this is me," were there other checks that were already in place that you didn't have to deal with? There was already a check with HR or the person in payroll in terms of onboarding these people. You didn't have to go and say, "Can you..." Because I was the same way, to be honest with you. I would find out when I was an admin, "Oh, so-and-so needs a Salesforce license," two days after they started when HR sent out the announcement of the new hires. And I was like, "Hey, there has to be some process that you onboard these people to the company. Can I be part of that?" And I kind of had to unwind HR and figure that out because they didn't know that Salesforce was a part of this department and they didn't know that people need to be onboarded for it. You sound like you didn't have to deal with any of that. Michelle Wolf: We did not. That onboarding process was already handled by a team. And so when the hiring manager said, "Yes, we're going to offer," and they accept, they already knew what systems, what equipment they need, what systems they need access to, and then what types of permissions they would need to ask for from Salesforce. So like what team they were going to, so that that request could be put in. All that hard legwork was already done. So to those who might not have that, I think the biggest selling point could be onboarding experience. A lot of companies are focusing on what that hiring process, what that onboarding experience is. And as someone's starting a new job, I want to prove myself early on. I want to say, "I'm here, I'm ready to work." And if I can't access my systems, even though it's fully out of my control, I still feel like I'm not doing my best as an employee, as a new employee. So you can leverage just even the onboarding experience that they have, the equipment they need, they have the logins that they need on day one, on the day that they need it. Mike: Yeah. And you don't have to go into detail, but what are the key, I guess I'll call them, milestones in your onboarding flow that the new people get sent? Do you have training materials, or do you have a video that's Michelle being like, "Hi, I'm your Salesforce administrator"? Michelle Wolf: If only. Mike: I know. I always say that and everybody's like, "Oh, that'd be a great idea." Michelle Wolf: It would be. It would be. Mike: "It's just 4,792 on my list of things to do." Michelle Wolf: Exactly. When I have that free minute, I will absolutely get to that next. Mike: Yeah. And also my other two admins, because it can't just be me. Michelle Wolf: Right. "Here's your Salesforce team. We're here to support you." Mike: Yep. Michelle Wolf: So what we built into our process was, in my Flow automation, there was actually a second autolaunched flow that would close out the case. And this did a couple of things. One, it closed out the case so they had a complete close of the loop and updated the reason as completed by automation because a human didn't have to touch any part of this request. Mike: Ooh, that's cool. Michelle Wolf: Yeah. Mike: I like that. Michelle Wolf: Clean reporting. Who doesn't like that? Mike: I mean, for a while, I remember doing approvals or something and somebody asked me, "Well, did you actually do this or did Salesforce, the Flow do it?" And I was like, "Well, the Flow did, but the Flow ran as me." Then it marks that. But I like that you added that completed with automation. Michelle Wolf: Yeah. But we also added two email actions. And one was to the requester. So the individual that opened the case through the screen flow, it sent them an email telling them that the request had been completed. And then the second thing was, when the user record was created, it then updated the Requested For field, which was just a custom field we made on the case for a user lookup with the new user that we'd made. And we sent a welcome email to them and said, "Welcome to Salesforce. Welcome to our company. Here's how you access our org." And it had the link for the SSO. Yeah. Mike: Because at the point that the person's creating the case for you to create the user, they've already had an email assigned to them. Michelle Wolf: Correct. Mike: I like it. That's pretty sweet. Michelle Wolf: Yeah. It was just a nice clean way to close multiple loops that ended up getting started. Mike: So then, not to dig into it, and this is where I wish I should have seen your presentation, do you also have a process for adding them to... I don't know if you guys have Slack or different things like that. How, outside of maybe that user emailing you, do they have not necessarily an open loop, but at least a channel to ping back the admins and say, "Oh, hey, I got everything except this doesn't work, or, "I'm not seeing X," or, "How do I upload a profile picture?"? Michelle Wolf: Yeah, absolutely. So for our new hires, basically all of that was handled in their new hire training, a lot of that initial setup support. However, in our screen flow, we actually made it a full-service utility where when you went to launch the screen flow, you could say, "I need help with Salesforce. Something's broken." Or, "I need a new user." So it was this full access path to us where they could put in multiple requests depending on what they needed. Mike: I like that. And I'm assuming it's not Michelle, but you have people in the organization that do training. How much do you have to keep them updated in terms of, "Here's what the new user experience is like for Salesforce"? Michelle Wolf: As far as? Mike: Just anything. Anything that would change or anything that they may encounter, new features. Michelle Wolf: Oh, yeah. So anytime we would build any type of enhancement, we worked very closely with our training department. We knew if we're going to help make starter material for an announcement for the call center for X, Y, Z enhancement, training's also going to need that material and probably a deep dive hands-on demo so that they can incorporate it into their training materials. So any new builds like that, our training department was really involved even from early-on stages of the development because they interact with new employees, and they see that interaction differently than someone who's been here for five, 10, 12 years and knows how we use our Salesforce. Even just how we design and where we put things on the Lightning page, they would often give us input from a new hire perspective to ensure that we're building good things. Mike: Now, just because we're nerdy here, did you have a way of capturing that? Was that a case as well? Michelle Wolf: That was just conversations. Mike: Oh, okay. Michelle Wolf: Yeah. It was, "Hey, we have this thing. We want a demo for you. Give us feedback," type of stuff. Mike: I didn't know if you were capturing that in terms of tech debt or things we need to build, things we'd like to build, and things we need more money to build. Michelle Wolf: Yeah, depending on what it was. It was just good old-fashioned scope creep. Mike: Okay. Yeah, that usually happens, unfortunately. Michelle Wolf: Yeah. Mike: I've talked about Flow and AgentForce. I've been around now. I remember from...
/episode/index/show/buttonclickadmin2/id/42332050
info_outline
Solving Sharing Mysteries with Setup with Agentforce
07/30/2026
Solving Sharing Mysteries with Setup with Agentforce
Today on the Salesforce Admins Podcast, we talk to Nikita Kothari, Senior Member of the Technical Staff at Salesforce. Join us as we chat about using Setup with Agentforce to understand and manage record access. You should for the full episode, but here are a few takeaways from our conversation with Nikita Kothari. Setup with Agentforce simplifies sharing Why can’t I see this record? It’s a simple question but, as any admin knows, finding an answer can get complicated quickly. If you’ve ever found yourself digging through Setup pages and running SOQL queries to troubleshoot permissions and record access issues, this episode is for you. This week, I’m talking to Nikita Kothari, a Senior Member of the Technical Staff at Salesforce. She’s here to tell us how Setup with Agentforce can help solve sharing mysteries. Using natural language questions, admins can trace access across org-wide defaults, role hierarchies, sharing rules, groups, and manual shares to figure out what’s really going on. Troubleshooting permissions and record access issues As Nikita explains, permissions and record access issues are so complicated because they are affected by many overlapping configuration settings. Small changes can accumulate over time, especially in a reorg. Setup with Agentforce was built to help you detangle these issues with simple natural language prompts. You can put your questions about permissions and record access to an agent, instead of having to wade through 1,300 pages of Setup to figure it out on your own. Once you’ve got a handle on what the problem is, you can use Setup with Agentforce to help you implement changes and get everything sorted. And it’s built with trust in mind: every write action requires your explicit approval, the agent is bounded by your permissions, and every configuration change is captured in the Setup Audit Trail. Plan permissions for your org If you’re trying to figure out where to get started, Nikita recommends starting small. “It's very difficult to go back and fix something,” she says, “so I would highly recommend trying any changes in a sandbox with the minimum set of users to see if it is working as expected or not.” And because Setup with Agentforce makes looking at permissions and record access so much easier, Nikita recommends taking advantage of it to conduct a monthly permissions review. Again, these issues accumulate over time, so an ounce of prevention is truly worth a pound of cure. Make sure to listen to my full conversation with Nikita for more on Setup with Agentforce and how to get sharing straightened out in your org. And don’t forget to subscribe for more episodes of the Salesforce Admins Podcast. Podcast swag Learn more Salesforce Admins Blog Post: Salesforce Admins Blog Post: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: Today on the Salesforce Admins Podcast, we're talking with Nikita Kothari, Senior Member of the Technical Staff here at Salesforce, about using Setup with Agentforce to understand and manage record access. Sharing can be one of those invisible parts of the platform until someone can see a record they shouldn't or can't see one they need to. Nikita explains how admins can use natural language questions to trace access across org-wide defaults, role hierarchies, sharing, rules, groups, and manual shares. We also discuss why validation, permissions, testing, and human approval remain essential when AI helps make configuration changes. So give this episode a listen, click that subscribe button, maybe share it with another Salesforce admin who's ever asked, "Why can this user see this record?" Let's get Nikita on the podcast. So Nikita, welcome to the podcast. Nikita Kothari: Thank you, Mike. It is amazing to be here. I'm from the engineering team, so you'll get my lot of perspective on how we are building Setup with Agentforce on sharing. Mike Gerholdt: Yeah, that's exactly what I wanted to talk about. Let's first get a little acquainted with you. What do you do at Salesforce, and what are some of the things that you've worked on? Nikita Kothari: So it's been close to two years I'm working with Salesforce. And currently, I'm working with access control and sharing space. What pulled me in is the realization that sharing is one of those invisible infrastructure layer. When it works, nobody thinks about it, but when it breaks, it actually breaks the customer trust. And you know how customer-centric Salesforce is, its number one priority is trust. So there is lot of critical issues that we are dealing with and we are trying to make the sharing as the best place for the admins. And we love our admins and we are trying to make things better for them. Mike Gerholdt: I love our admins too. I used to be one for a while. I'd like to still think I am, but we'll see. Well, let's get into it. So I read your article and I'll link to that in the show notes. Tell me, I mean, before Setup with Agentforce, what did an admin have to do to answer a question of, why can this user see this opportunity and this user not? Nikita Kothari: Oh, good question. So it's the simplest way of dealing with solving the sharing's toughest problems. So just to enable the Agentforce, you don't have to do anything. It comes along with the Agentforce. So go to the setup page, search for the Agentforce agent and enable your Setup with Agentforce, and then you can start writing those questions. You don't have to know anything beforehand, that's the best part of the Agentforce setup. And sharing is the challenging domain, I won't deny that. It is very, very complex. So answering even the simplest question that you ask, like, "Who this person is having access for this thing," it requires to look into a lot of different domains like OWD, role hierarchy, sharing rules, sharing groups, manual shares, to see which layer actually give that access to that particular user. And looking that information, sometimes it's easier with the UI, but most of the time user has to run lots of SQL queries to get the correct answer. And previously, admin was spending hours of work and sometimes even the afternoons just to deal with that single questions. And with Setup with Agentforce, it just a minute of work. You just give prompt to your chatbots and it'll give you the correct answer within seconds. Mike Gerholdt: Yeah. I mean, boy, it used to be a lot of archeological digging. I think you called it that in the article. I know I used to have to feel like I was going through a checkbox, but tell me, I mean, walk us through a real investigation where an admin discovers that a user can see a record, they probably shouldn't. How would they use Setup with Agentforce to trace where that access came from? Nikita Kothari: Yeah, great question. So while building these actions, we particularly divide it into two different segments. First segment is the read action and the second segment is the write action. Just to keep it, things untangled, read actions will help you to investigate. When you ask a question like, "Why Nikita has access for these things?" So our agent will go into the background, it will do all kinds of research and it will come up with the best answer. "Okay, Nikita is part of this group, and these groups provide the access to this particular record." And what a write action on the hand will do, it will help you to solve that problem. Now you know Nikita is part of the particular group. You can go ahead and ask your write action, "Can you please remove Nikita from this group?" And your write action will help you to remove that particular user from that group and boom, your problem is solved within a few minutes. It's reliable, it's fast, and I'm sure admin can save a lot of time with these two things. And you don't need to know which action or which thing you need. You just write the plain natural language and our LLM will be able to guess where to navigate and how to solve that problem. Mike Gerholdt: Yeah. No, I hear you. I think you walked us through what a good sharing investigation looks like. But for you, is there a sequence admins should follow so that they don't just jump straight to changing configuration? Nikita Kothari: I would say I would highly recommend admin to start playing with the read action. There is no particular sequence, I would say, but mostly how the sharing is defined is like first we have the OWD, which provides the base access to the records and to the objects. And then comes the role hierarchy, then comes sharing rules, then we have public groups and queue and we have the manual shares. So at each level, sharing opens up the door for more access. So if you're going below the hierarchy, sharing will provide more and more access to the user. And if you want to close that access, it's not easier that one sharing rule will open up the access and then you create another sharing rule to close the access. It's not how it works. So in that case, you need to spend a strategic amount of time thinking how you can revoke that access. And sharing is accumulation of the small, small, small decisions over the time. Sometimes adding someone to the group, for the reorg, you have to do lot of things with the sharing, and every decision gets compound. So I highly recommend our admins to at least spend 30 minutes a month to refresh and reanalyze their sharing configuration to avoid the future problem. So this is basically the flow looks like, but I would say there is any of sequencing which admins should follow. It's pretty much open and admin can ask any question to our agent. Mike Gerholdt: No, I think you're right. I mean, Setup with Agentforce can do a lot in terms of explaining the configuration and then help change it. I think you really touched on that last answer of just slowing down for a second, verifying what the answer is to make sure that you're staying in control. My question to you would be, how should an admin validate what the agent tells them? Nikita Kothari: Oh, that's a good question. So for the read action, the information, we have mostly the links added to our prompts. We call it as a clickable link. So if investigation is saying that Nikita is part of this group, actually, there is a link of that group. So you can go to that group by clicking on that link, and you can actually see I'm a part of that group if you want to re-verify the stuff. So, that will give you more confidence that you know how admin is investigating the information or if admin is saying, "Nikita is part of the sharing rule and you'll have the link for the sharing rule." So you can actually go and verify whether that sharing rule contains the group or the users where it is providing that access. And with the write action, I think more critical is the right action. I mean, you don't want somebody, AI is coming to your system and randomly making changes to your system. So to do that, we have the validation layers. The first validation we generally come up with is for every write action, we explicitly ask for the approval, approval from the person who is making the changes. And we also have the permission boundaries, and if particular user doesn't have the permission to edit OWD, the agent won't be able to edit OWD for that person. And then the last part is the accountability. Every change we are logging into the setup audit trail saying that whether the AI agent did it or actual person did it. And then admin can verify it, whatever changes is made by the AI agent are trustable or not. So we will always advise admin to go and do as much as possible testing on the sandbox before making changes into the production. And another backup system is the Einstein Trust layer. We always say trust is our number one priority. So, Salesforce has built trust everywhere to make sure that we are not blindly making any changes, and as this is a very, very critical domain to our customers. Mike Gerholdt: No, that's great. You talked a lot about sharing, and I always feel like making a change with sharing, you want to make sure that you're doing it right. In your opinion, what is the most important thing an admin should review before making a sharing change? Nikita Kothari: So I would say, as I said before, we have some of the steps that I also wrote in my blog, that we have OWD sharing rule, rule hierarchy, and everything opens up the more access. So to close that access, it's very difficult to go back and fix something. So, I would highly recommend whenever our admins are making the changes, try it out first thing on the sandbox if it is working as expected or not. Or have some verification system or another set of eyes whenever you create some rule hierarchy or you're providing some access to the rule, rather than giving access to all of the user, first try with the minimal set of the user. Try with the one user and see if that user are getting right access to the right record. Because even if you fail to provide a single user a wrong access, that means that you're leaking your data. So having those, another, just starting with a small and then going into the bigger and bigger data, that would really help admin to build the reliable system. Mike Gerholdt: Yeah, absolutely. I think it can do a lot and I've seen it do a lot, but what is something that Setup with Agentforce can help with today that maybe admins may not realize? Nikita Kothari: I feel like still adoption, it's tricky because people are quite a lot thinking that Setup with Agentforce will replace their job. But I would say rather than replacing, it will provide you a lot of capabilities and it will remove a lot of overhead from your day-to-day work life, rather than going and looking into the SQLs or doing everything by yourself, just give it a try. And then you can see how much opportunity it is opening up for you to save time. And that time you can actually use for building a more structural, and because structure mattered a lot when it comes for sharing. How your roles are getting access or how your groups are getting access. And if you're just blindly following the steps and not thinking much, then you are in a trouble. So you can spend more time and thinking about, how can you better make a structure for your organization? And you can get more time to think about your business. And I feel like most of the troubleshooting cases we have covered into the Setup with Agentforce flow. Yeah. Mike Gerholdt: Yeah. You mentioned a lot, admins should get out and use it and try it, and I'm a big proponent of that. So I'll end on this question. What for you is one kind of low risk question that an admin could ask Setup with Agentforce today to better understand their sharing configuration? Nikita Kothari: I think my one of the favorite action among all of it is record access, to ask why this person is having access to this record. And you see how much it can do, it will give you the whole list of the OWD configuration for that particular record, or it will look for the sharing role, it will look for the group membership, it will look for the manual shares. And it will do all kind of investigation for you just when you type a one single natural language to your agent. So you can always think about someone which is having more knowledge about the setup. It's sitting with you and helping you to deal with all of the setup configuration. Because setup, again, we have like 1,300 setup pages, which is lot. And going through all and learning the capabilities of each setup action, it's difficult. So, that's why we built the Setup with Agentforce, which will give you the capability that you don't have to learn everything on the go. And you can start asking question to our bot. If it won't be able to answer, it will guide you in the right direction to help and analyze your queries or your concerns. Mike Gerholdt: Yeah. I am so glad that we have Setup with Agentforce. I feel like I could have used it 20 years ago. I know I had a lot of questions and would have to dig through my notes on org-wide and sharing and group sharing and all kinds of that stuff. So, I appreciate it. And I appreciate the article and you taking time out of your day to talk to admins and share your knowledge about sharing and setup and how we can do things better. Nikita Kothari: Yeah. I think admins are always on the top of our list. I would love to hear from our admins how they like this Setup with Agentforce, specifically with the sharing, how it is helping them or what they are looking forward to, because I feel like most of our critical projects and best solution came up with the admin feedback. And we really want to make admin life easier. We always have meetings internally, our engineering team, our product is always so much focused about thinking about our admins, thinking about the newer way to introduce some more features to make their lives easier so that they can focus on building a reliable solution. So, it was great talking to you. Mike Gerholdt: Yeah, no, I'm glad you could share your knowledge with us today. Thanks so much. Nikita Kothari: Yeah, thank you so much. And one more thing, we will be at Dreamforce. Larry Sherrill is our Product Manager. They're having a lot to share at Dreamforce. So, feel free to stop at user access booth. I'll be there too, so come and say hi and share your feedback. We would love to hear back from you. And we keep making a lot of progress on Setup with Agentforce, so stay tuned, follow our blogs, and it will be something will be really helpful in future for you guys. Thank you so much. Mike Gerholdt: Thanks again to Nikita for helping us understand how Setup with Agentforce can make sharing investigations faster without taking judgment out of the Salesforce admin's hand. And the takeaway is just really simple, use AI to uncover the source of access, validate what it tells you, and test changes carefully before they reach production. Now, if you enjoyed this episode, and I bet you did, I hope you did, you should subscribe to the Salesforce Admins Podcast. Share this episode maybe with your team, fellow team, user group. Oh, I like that, if you'd share it with the user group, that would be awesome. And then, how about this? Start with one low risk question about your own org sharing configuration and ask that to Setup with Agentforce. Until next time, we'll see you in the cloud.
/episode/index/show/buttonclickadmin2/id/42257655
info_outline
Salesforce Superbadge Cohorts Build Skills Through Community
07/23/2026
Salesforce Superbadge Cohorts Build Skills Through Community
Today on the Salesforce Admins Podcast, we talk to Jean Velonis, Senior Technical Program Manager at Salesforce. Join us as we chat about the new Superbadge Cohort Learning pilot program and how you can get involved. You should for the full episode, but here are a few takeaways from our conversation with Jean Velonis. Improvements to Superbadges I have to confess, I’ve always found Superbadges to be a little intimidating. They're locked behind requirements, without clear step-by-step instructions for how to work towards them. That’s why I sat down with Jean Velonis, a Senior Technical Program Manager working to improve the Superbadge experience. The first thing Jean wants you to know is that they’ve overhauled the Superbadge UI. Instead of being locked behind a list of requirements, Superbadges now have a Recommended Learning section. These Trailhead modules and badges are organized into units, giving you a much clearer path to work towards your goal. What is the Superbadge Cohort Learning pilot program? In addition to the UI changes, Jean and her team are launching a new pilot program called Superbadge Cohort Learning. These sessions let you connect with other learners and subject matter experts to work towards a Superbadge together, guided by a facilitator. For now, the pilot program is focused on two of the most popular Superbadges. For admins, the Data Quality and Validation Superbadge gives you the knowledge you need to prep your org for AI. And if you’re a dev or looking to improve your coding skills, there will also be groups for the Apex Callouts Superbadge. They’re running both in-person and virtual versions of these sessions, so look out for them at a Dreamin’ event or community group near you. Volunteer as a subject matter expert or facilitator If you already have these Superbadges, Jean highly recommends volunteering as a subject matter expert or a facilitator. When you’re working with new learners, you’ll run into questions and situations that deepen your understanding of the topic more than you ever thought possible. The other thing about students is that they keep learning. Teaching is an incredible opportunity to make strong connections that will last throughout your career. As Jean explains, two of her former students are now her go-to people when she has code or Flow questions. Make sure to listen to the full episode for more from Jean about the Superbadge Cohort Learning pilot program. And make sure you’re subscribed to the Salesforce Admins Podcast to catch us in your feed every Thursday. Podcast swag Learn more Sign up to be a subject matter expert or facilitator: Salesforce 360 Blog Post: Admin Trailblazers Group Social Full Transcript Mike: This week on the Salesforce Admins Podcast, I'm joined by Jean Velonis to talk about the Superbadge Cohort Learning Pilot and what happens when admins stop learning alone and start solving problems together. We're going to dig into these cohorts and how they bring experts, beginners, and everyone together in the same room to work through real challenges, build confidence, and get comfortable being uncomfortable. Because for Salesforce admins, knowing how to troubleshoot, test, fail safely, and ask better questions matters just as much as knowing where to click. And as AI and agents become part of everyday platform work, well, those fundamental skills help admins validate what is correct, keep data reliable, and make sure that the platform reflects what the business actually needs. So listen in, subscribe and share this episode with an admin who learns best by rolling up their sleeves and figuring it out. Let's get Jean on the podcast. So Jean, welcome to the podcast. Jean Velonis: Thank you so much for having me. I think this is the second time I've been here, but it's been a while. Mike: I know. Well, when you do the podcast for almost 14 years now, I try not to have guests back, only if they're important like yourself. Jean Velonis: Ooh, thank you. Mike: We were Salesforce MVPs together a hundred years ago when the internet was still tubes and wires, and now we're both at Salesforce. And you're on, I would call, what, the curriculum side, the learning side? Jean Velonis: Yeah, I would categorize that. It's more Trailhead, right? So long story short, certification was my jam for so long, and superbadges were part of that. But over the last couple of years, we've decoupled from certification. We're still with Trailhead and we're still considered an assessment, but we don't have all the guardrails of certification on us anymore, which is really exciting. Mike: Gotcha. And just for reference, if anybody thinks they might have met Jean at a Dreaming event or a Salesforce event, you tend to usually have purple hair and a bedazzled Ranger hat. Is that an accurate description? Jean Velonis: These are facts. These are facts. I've had purple hair probably since six months of coming into Salesforce. And if anybody knows the famous Chris Duarte out there, she taught me everything I know about bedazzling. So I made a Ranger hat. Mike: Yeah. A lot of people bedazzle now because of her. Jean Velonis: Yes. Mike: Well, let's talk about this new Superbadge cohort learning pilot that you've kicked off because I think you're trying to bring the best of both worlds together. Jean Velonis: Absolutely. This isn't a new concept either. I'll just throw that out there. We had some amazing trailblazers before the pandemic and even through the pandemic doing Salesforce Saturdays or Salesforce Sundays and they would get together in person or virtually and work on Trailhead. So this is kind of piggybacking off that concept and really bringing some of the community magic to superbadges because I think we can all look at a superbadge and be like, oh, that's kind of intimidating. There's no step-by-step instructions. This looks really hard. There's multiple steps to it, and that just looks hard and I don't want to do it. But the cohort learning brings a facilitator, like a community group leader or a Dreamin event person to come in and help bring people together virtually or in person and work on a Superbadge together. And we've been running a couple of pilots of these. We did one at TDX and WITness Success in Indi. And it's been really successful in person just because you can pair experts and newbies together at a table and make people really uncomfortable, and then they start talking to each other and learning from each other. And I think that's what gets really excited about doing this cohort learning. And then we've done a couple of virtual events as well where we've brought community group leaders, kind of like train the trainer of like, here's how you should run this virtually to be successful, because that virtual is a little bit different beast. It's really easy to hide and stay on mute and turn your camera off where we really try and get people excited to share their skillset with each other. Mike: Yeah. No hiding in the dark corners. So I really like the idea of this. I've always really enjoyed Trailhead because I'm the type of learner that likes to do it at my own pace. Sometimes I even crazily go back and redo a badge just because I feel like if I've rushed through it, did I really learn it or was I just trying to get it done? And when I'm really trying to learn it, I really take my time. I will agree with you, I've written superbadges. I also enjoy the freedom. They're kind of like, "Here, go, do. " And you're like, "Oh yes." It's almost like the days of driver's ed when you get out from behind the desk and you finally get to drive the car in the real world. It's a little more of trying things out. So are these cohorts that you're doing now, are they open to any superbadge or are they just specific ones? Jean Velonis: That's a really good question. We looked at our whole library of Superbadges and we wanted to pick one for admins and one for developers. And that doesn't mean it's just if you have that skill set. It's more of like, here's two of the main roles we see in the community. So we have data quality and validation, which is really for anybody prepping you for this agentic ecosystem that we're in, cleaning up your org. Mike: I mean, to be honest with you, that's what we've needed to do this whole time. Even before AI, it's sweeping the floors, you know? Jean Velonis: Yeah. I mean, it's been that way for a long time. I think I made my whole Salesforce admin career of going in and cleaning up orgs, to be honest. So that was a great place for us to start. And then now we've started looking at the developer side of it and we picked Apex Callouts superbadge for all of our developers. And I was very intimidated by trying to put together a superbadge in a box for that cohort because I'm not a developer. But going through this process, I'm like, "You know what? I think I can actually do this," And that's because of all the great learning badges on Trailhead and being able to go at your own pace. Because that's a whole nother skill set for an admin to try and figure out, right? Mike: Right. I've often thought of, this is just how Mike's brain works. If I had children, I would for sure name one of them Apex because then you just gave them a shout-out. Jean Velonis: That's hilarious. I would've never thought- Mike: That's how I work. My brain today is like, "Oh yeah. Hey, did you hear you were mentioned on a Salesforce podcast?" "Of course I was. I'm an Apex Callout." Jean Velonis: I'll tell you a weird secret. I do put my children's names in superbadge content. All of us over here under the superbadge team put little- Mike: Easter eggs? Jean Velonis: Easter eggs, yeah- Mike: Yeah, I love it. Jean Velonis: ... out there. It's either in the org or it's in the content. We're all moms over here, so we would like to give our kids a shout-out. Mike: Well, I will see your Easter egg and I will call your Easter egg. In almost every admin keynote demo, we have put our pets' names. Jean Velonis: I love that. Mike: So you can go back and figure out whose pets were whom. The only time, the only exception was we had Einstein Voice rollout and my dog's name was yet to be used and we couldn't use it because it wouldn't recognize it. It was too hard to say. And so we had to go with something easy and punchy with hard consonants in it. So anyway, yes. I think there should be, I'm sure there is somewhere, a whole forum of Easter eggs found in Trailhead material. Jean Velonis: 100%. And I think that's what Trailhead is, where it's the weird, it's the hanging out, it's the learning something new, it's the being uncomfortable. Because it really is, learning can be really uncomfortable. Mike: I mean, it exposes a weakness and people don't like feeling weak. It's also the thing I always think about whenever I was doing training because you're changing something. And the person that told me how to do training gave me a very well-rounded, concise piece of advice, which is people don't like to go to work and feel stupid. And when you change something, even in a CRM, that's what that comes from. And people don't like to feel stupid. And I feel that way. You change something, you're like, "Oh, this was here yesterday. Why isn't it here?" And you blame yourself and that's not the point. You do a lot of training. So let's go through just some scenarios of I'm a new Salesforce admin. I stumbled across this podcast. Also, hello. You have a lot of episodes to catch up on. That's okay. You have a lot of time. What should they do to prepare to come to one of these learning cohorts? Jean Velonis: I mean, the easy answer is get on Trailhead and start earning badges, that's number one. Step two is take a look at the new look and feel for superbadges and that will bring you to our recommended learning. It used to be required to unlock the superbadge, but we've just now come onto what I would say the Trailhead UI. Superbadges are put into these units now where the first unit is like, let's get ready. Let's get your developer org. Let's do the recommended learning. Let's connect it to your superbadge now. That's all a little bit different than what we used to do with superbadges. You had to unlock it and then you had to go down all the way to the bottom of the page to hook up your org. Now it looks like a regular module or badge, which is nice because as somebody who struggles with reading and keeping my place on Trailhead, this new guy is really going to help with that. But also doing the recommended learning I think is key even though it's unlocked now and you don't have to do it. I will tell you I am the first person to start doing something and being like, "Oh, I have no idea what I'm doing with this. Maybe I should have done-" Mike: I should have done those recommended modules. Jean Velonis: Yes. And then I will have to go back and learn and figure things out. I did that with the prompt template superbadge. I thought I knew what I was doing. Mike: No. Jean Velonis: I did not. And that's falling, that's failing forward. That's realizing I don't know what I'm doing. Let me go figure this out and then come back to it. So I think that's step number two. Go and read through the superbadges, look at them, look at the recommended learning. And then if you see that there is a Dreamin event or a community group doing one of these superbadges, join them. Even if you've done the superbadge, that's okay. You can go help the other people that are there. Mike: Oh, I almost think that would be the next best thing after doing a superbadge is going and volunteering to help train. Because never will you ever learn more about something than when you have to train somebody about something or walk around the room and answer questions. You're like, "Hey, I didn't run into that when I was doing mine, but let's figure it out." And it's super fun because then you're like, "This wasn't a problem for me, but now it's a problem that I get to work through." Jean Velonis: Yeah. I mean, bringing you all the way back to SABWA. Mike: Oh, sure. Jean Velonis: Call out, Mike. I mean, geez. Yeah, that is my favorite thing to walk around, especially I live in these superbadges every day. Our team builds them. We maintain them. We have to deal with all the product changes that are very quickly changing and trying to keep up with everything. So being able to walk around the room and see how trailblazers are doing things, that not only helps me figure out, okay, here's some of the common pitfalls or where people are getting stuck and I can write some hints to help, but also maybe we don't do that in the next superbadge or maybe we call it out in a different way or let's go update the help article and call this piece out if somebody's getting stuck somewhere. So it's super helpful. But also you get to meet a ton of different people, which we do virtually every day, but it's easy to hide behind the keyboard again. And to be able to connect with somebody and see their aha moment of where they're like, "Yes, I figured this out," that's super gratifying. Mike: Oh, that's almost the time that you make a friend. Jean Velonis: A best friend. Mike: Yeah. I mean, because you're always going to be like, "Well, let's continue chatting while you work through this." And then the next thing you know, they're building Apex Callouts, and you're like, "Oh, I didn't get that far." And then they're teaching you stuff. I promise you it'll happen. Jean Velonis: It will happen. I can say 100% certainty it will happen. I have two people that I call on all the time or that I met through a conference or doing something, and I'm like, "Okay, I know this person knows code and I know this person knows flow. I'm not good at those things, but I can text or call or Slack and be like, I'm stuck." Mike: Right. So I mean, I have a thousand questions, but let's go to the opposite end of that where if you're a community member, maybe you're a user group leader or a Dreamin event person and you're like, "This sounds like a super fun thing for me to include in my agenda," What would their steps be for possibly including, I know it's in a pilot, but for including this or what's your vision for it moving forward? Jean Velonis: Yeah. We have a facilitator signup forum- Mike: Sweet. Jean Velonis: ... which I'm sure you would be happy to put that in the bio or the show notes. Mike: It is in the show notes as we speak. Jean Velonis: Perfect. But yeah, just signing up to raise your hand to be a facilitator is huge. Because I'm going to have this superbadge in a box ready very quickly for people to take to a pilot. I have two Dreamin events that are already wanting to do this in August and a community group that also wants to bring it to their local people. So signing up to be a facilitator and doing this on your own, love that. And I wish I could go to every one of them, to be honest. I looked at a couple of them and I'm like, "I might be able to make this work and just show up." But there's also a piece in that form too to sign up to be a subject matter expert and help if there's somebody locally that wants to do this. And selfishly, I put that in there that if you want to help us build superbadges, that would be awesome too. Mike: Right. And this sounds rhetorical, but it is honest. They should have probably already completed that superbadge that they want to facilitate or host. Jean Velonis: I mean, I'd advise it. And with the new UI, you can redo a superbadge. You can retake it now. So you can do the superbadge over. Mike: To refresher, because maybe you haven't done it in a while and want to see what's new. Jean Velonis: I mean, the facilitator that did WITness Success told me she still had her notes when she originally did the Apex Callouts in 2017. Mike: Wow. Jean Velonis: Right? Mike: I probably have notes from somewhere of 2017 I have nowhere. That's awesome. Jean Velonis: No, that notebook's gone. Mike: Yeah. I always think of admins listening to this and I don't want it to be a pitch session for your cohort, but this is really a neat way to get involved with the community. From all of your work, and this is just stepping out of your role and what you're doing in this superbadge cohort. Out of all of the work that you've done at Salesforce and what you create for Trailhead, what are some of the biggest themes that you run into that really, when that light bulb turns on for an admin, they suddenly just become, I don't want to say instantly become successful, but they really start to click and feel confidence and have that aha moment? Jean Velonis: I think being a Salesforce admin for such a long time, being able to fail and problem solve is probably the best skill you can have. You can learn any technical skill. I can go into a dev org and break things and figure out how products work. But if you don't know how to problem solve through something, getting to that aha moment is going to take you a long time if you're too afraid to break something. And I have broken things. And I think I've told you, I've deleted records from an org before and went, "Uh-oh, how do I get those back?" So that is one of my favorite things is to go into a playground or dev org and just break it, pull things apart, see how it works and be like, "Okay, I know where I want to go, but how do I get there? And I don't know how to get there, but if I go in and break everything, I'm going to learn a lot more than if I follow step-by-step instructions of how to do something." Mike: Right. If it works out perfect the first time. Jean Velonis: Yeah. What did you learn? Oh, I know how to do that. Okay. Well, what's next then? Mike: Do you find that you have to remind people that it's okay to fail when building and it's okay for things to not work the way you want them to the first time? Jean Velonis: Yeah. And I mean, that's how we build superbadges. I mean,...
/episode/index/show/buttonclickadmin2/id/42185255
info_outline
How Can Admins Reduce MFA Friction in Salesforce?
07/16/2026
How Can Admins Reduce MFA Friction in Salesforce?
Today on the Salesforce Admins Podcast, we talk to Jay Hurst, Senior Vice President of Product Management, and James Ferguson, Senior Director of Product Management, at Salesforce. Join us as we chat about MFA step-up authentication and what it means for Salesforce Admins. You should for the full episode, but here are a few takeaways from our conversation with Jay Hurst and James Ferguson. Step-up authentication protects sensitive actions Starting next week, Salesforce is requiring all users to use Multi-Factor Authentication (MFA). If you’re a privileged user like, for example, an admin, you’ll need to use a phishing-resistant MFA. That’s why I sat down with Jay Hurst, VP of Product Management, and James Ferguson, Senior Director of Product Management, to talk about why these changes are vital to protect your org’s data. The first thing to know is that AI is making it easier than ever to launch targeted phishing attacks at scale. So while the MFA requirements provide a good first layer of protection, we want to make extra sure you are who you say you are before you’re allowed to perform certain actions, like downloading a large number of records or running a big report. Phishing-resistant MFA uses a passkey, like a fingerprint or facial recognition biometric, to verify that it’s actually you and not just someone with access to your email account. Balancing security with user friction As Jay and James acknowledge, these changes will add some friction to your users’ experience. However, with the pace at which these kinds of attacks are evolving, it’s more important than ever to get serious about your security posture. “We're trying to introduce a little more friction right now so that people start to think,” Jay explains, “and start to build those habits of understanding when they're doing something that potentially could be considered a malicious attack, such as downloading that All Opportunities report.” They’re also building out compensating controls that should make things easier in the future, allowing you to trust users from a certain IP range, for example. Security is a journey, not a destination The most important thing to realize is that these requirements are about more than just jumping through some extra hoops. Phishing and man-in-the-middle attacks are growing more and more sophisticated, and you need better protections than “Well, that hasn’t happened yet.” Instead, James and Jay recommend viewing this as an opportunity to partner with other stakeholders in your org to develop a comprehensive security plan. As Jay says, “Security is a journey, not a destination. What is 100% secure today is not as secure tomorrow.” The trick is to develop a security-focused mindset throughout your business that will protect you now and in the future. Make sure to listen to my full conversation with Jay and James for more on step-up authentication and how admins can reduce friction for users. And make sure you’re subscribed to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Podcast Episode: Salesforce Admins Blog Post: Salesforce Help Article: Salesforce Help Article: Salesforce Help Article: Salesforce Help Article: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins Podcast, we're talking with Jay Hurst and James Ferguson from Salesforce Product Management about MFA step-up authentication and what it means for Salesforce admins. As you know, security isn't just a front-door login decision anymore. It's about protecting sensitive actions, understanding risk, and designing systems users can trust. So Jay and James are going to help us unpack phishing-resistant MFA, compensating controls, IP ranges, SSO, and why these changes matter in a world where data, automation and AI are all working together. For us Salesforce admins, this is a chance to think beyond features and really look at how we steward the entire system. So listen in, click that Subscribe button, and of course I would love if you could share it with fellow Salesforce admins or, hey, you know what? Let's make some friends in that security team. So with that, let's get Jay and James on the podcast. So Jay and James, welcome to the podcast. Jay Hurst: Thanks for having us, Mike. James Ferguson: Great to be here. Mike: Absolutely. Jay, let's start off with you. We kind of want to get to know a little bit about you, and James, we'll call on you second, but before we get into our topic today, can you just tell me a little bit about how you got to Salesforce and what you do? Jay Hurst: Sure, yeah. So I have been with Salesforce for almost 22 years now. I started in our customer support department, one of the first 12 phone support reps here at Salesforce. Did that for a couple years and helped found our Tier 3 organization in support. Eventually moved over to our customer-centric engineering department, stayed in there for a while. And then in 2012, had an opportunity to join the product management group for platform, and I moved over and ran a team called Force.com Canvas. And for the last, I guess, 12-ish years now, I've been kind of weaving my way upwards through platform. Currently, I lead our platform services subcloud, so all of the core foundational pieces of platform that you might think of are schema and metadata, APIs, eventing systems, connectivity systems, and also our identity area, which is what brings us here today to talk about MFA. Mike: Yeah. Boy, flashback. You called it the Force.com platform. Jay Hurst: Well, that's what it was called back then. Mike: I know. I know. Jay Hurst: And I can't remember all of the names we've had for it. Mike: Oh, that's okay. I'm sure there's a website that tracks all of them. Jay Hurst: I'm sure there is. Mike: James, fill us in. How'd you get to Salesforce, and what do you do here? James Ferguson: Well, I am, I guess compared to Jay, one of the newer members of the team. I've only been at Salesforce for about 16 and a half years, almost 17 years. Pretty much entirely on the platform product management side, working on various things people know and love like sharing and big objects and event monitoring and those things. And most recently I've taken over responsibility for the identity product team, responsible for all the login and auth and SSO and all of the wonderful things we'll talk about today. Mike: Oh, wow. Okay. So then just to be clear, I'm actually the newest person on this call. I've only been at Salesforce for a little over 12 years now, so I guess I still have my rookie stripes. Jay Hurst: Combined we're almost at 40, or just over 40. Mike: Yeah, combined. We almost get our AARP discount, right? Jay Hurst: Yeah, exactly. Mike: Jay, let's kick off. I know I did a podcast ... and I'll link back to it ... not that long ago with Laura Pelke talking about some of the new things that were coming out, and of course security is always big on admins' mind. She did a wonderful job of explaining step-up authentication to me, which was basically the airport analogy of you have to show your ID to get in and then you have to show your boarding pass to get onto the plane. I thought that really made sense to me, but let's talk about the new authentication that's coming out, if you call it that, and the new step-up concerns that Salesforce admins have. Jay Hurst: Sure. So I think as we move into the continued proliferation of agents and AI across the industry, security is obviously top of mind for a lot of our customers and for Salesforce as well, specifically because we have to help protect our customers. And so when we're thinking of that and how we ensure our customers' data is protected, with step-up authentication, it's really focused around in that same analogy, making sure you're providing your boarding pass at the right times when you're doing things. So just like you need to show your boarding pass when you go through the TSA gate and when you're on the plane and probably to the gate agent after you're on the plane, when you're doing certain things within Salesforce, we want to make sure you are who you actually are and your session hasn't been compromised. So when you're doing certain higher sensitivity-type actions such as I want to download 10,000 records out of my system, maybe run a report, putting that end user through another verification of, "Hey, is this actually you? Prove it with your step-up," so that we have that confidence that we can release the records. And so this kind of helps prevent some of those man-in-the-middle phishing attacks where somebody gets you to log in and then steals your credentials or steals your session in the background. So it's kind of that second or third or fourth level of protection in the runtime. Mike: Wow. James Ferguson: I think that's an important shift that's worth calling out, because it's no longer about just putting a stronger lock on the front door and making sure somebody has better passwords, or even the later stuff, the more recent stuff with verification. But it's when sensitive things happen, we need to do a little extra even once you're inside the airport, if we want to continue that analogy. And so it's a shift from that front door to moving forward. Mike: Yeah. James, help me understand that a little bit more, because I think one of the things that admins always fight is user friction. How hard is it to do something? And now we're introducing something when it could disrupt their flow of work, but it's for a good reason. James Ferguson: Exactly. I would say that we are constantly balancing that. We're constantly balancing the need for security with the friction it does. And so we aren't expecting to do every click, for example, but just when you're exporting data from a report, which is potentially pulling a large amount of data out, we want to make extra sure. We're looking at some other things in the future around maybe when you're an admin changing some security configurations, we want to make sure you are who you are. And frankly, you see this more and more even in consumer websites where sometimes you're asked. If you're going to change a phone number, change a thing, you need to go back and sort of double-proof who you are. But it is a trade-off. It is a trade-off. Mike: Yeah. It's like once you're logged in, it's, "Oh, if you're going to change." I was trying to order tacos the other day through an app and I needed to update my credit card. And then right after I did that, they're like, "And you need to put your password back in." I'm assuming that was kind of the same situation. James Ferguson: Kind of the same situation. Jay Hurst: Yeah, exactly. Exactly. Yeah. And the other thing I'll add to this too is on the friction point. We always have to balance that. Like James said, we don't want people to become overly frustrated and not want to use Salesforce because it has too much friction. But the reality is over the last 20 years, we've edged so far on the line of not causing any friction that we've led admins down the path of not having that understanding of what their users are doing, and really not getting into that proper security-conscious mode. So we're trying to rubber band a little bit back on the other side, introduce a little more friction right now, so that people start to think and start to build those habits of understanding when they're doing something that potentially could be considered a malicious attack, such as downloading that all-opportunities report. So starting to get customers to think through that while we are continuing to build out what we call compensating control. So the end goal here is not to provide friction and it's not to force you to do these one-time passwords for everything you need to do in the system. It's to give admins that layered capability of saying, "Well, if I have these five or six different controls I can put into place, then they can kind of compensate for one another." So where the most friction might be having to open up your email to get that one-time password to do the step-up, maybe we can compensate that through things like having a trusted IP range or already having a phishing-resistant authentication put in, or other controls that we will layer in. So we ultimately get to a posture where the admin can both reduce friction while choosing the proper set of controls for their needs. Mike: Yeah. No, that makes sense. You mentioned phishing-resistant MFA. Can you help me explain that if somebody's never heard that before? Jay Hurst: Yeah. So maybe I'm going to try and wiggle this airport analogy in. If we think way back in the day when you had your IDs, before REAL IDs and all of this, you didn't actually have good authentication on them. You might have a really bad picture, or in some cases IDs didn't have any pictures on them. They just had your name and your information, but they were accepted. Now, the problem with that is anybody could steal it if it doesn't have your picture, and they can pretend to be you. So phishing-resistant MFA is kind of that same thing. Where a normal MFA might be a one-time password on your phone where you get that push notification that says, "Hey, your code is 123456, enter that code in," that's great because it means you have to have your phone in order to get that one-time password, but it's not as secure because it's sent over SMS. It doesn't really verify that you are who you say you are. It's only that you have access to that phone. Phishing-resistant adds another layer on top of that, so it's not just over SMS or in an app. It would be through what are called passkeys or other more cryptographically secure passes. So think of your face ID, your touch ID, some of those passkey-type implementations that you might see, where it's not just getting a code, but you have to unlock a vault that has your information in it and then you provide that information out. So it's taking it from that ID that anybody can use if they just happen to steal your wallet into more of a REAL ID that also has some biometrics attached to it. So you can actually guarantee it's not just that you have the thing, but you have the thing and you are the person that should have the thing that identifies you. And so that's what we're implementing today that's starting to roll out tomorrow in production. We've already started pushing it into sandbox for admins where we want all of our administrative users across the system to enroll and use these phishing-resistant MFA capabilities, so that they're more secure. Mike: Yeah. I never even thought of that. All the time there's apps I log into and they send me a one-time passcode and I'm like, "Well, of course I have access to it." Never did I think that, "Oh, what if somebody already has access to my email, requests that code, and now they're essentially spoofing me?" And it never dawned on me until you think of, well, they're not really verifying that it's you, Mike. They're just verifying that whoever requested that code also has access to your inbox. And you're like, "Oh." Jay Hurst: Exactly. And that may not be a nice person. Mike: Right. So James, what are some questions that admins are asking online that we can kind of help answer for them, or equip them for when they get this rolled out to them? James Ferguson: They're certainly asking questions about phishing-resistant MFA versus regular MFA and the different sort of authenticators that are out there. I'll shift this around a bit. One of the things they aren't asking about, it's actually something that Jay mentioned in passing, which is the alternate controls. And for things like step-up, if you have login IP ranges on the org or on the user profile, or there's some session settings which force you to always be coming from the same IP address in your session, which is the case for most of us who are at a laptop at work or whatever it is. Those avoid step-up entirely because we have more certainty about who you are and where you're coming from. If your session's bouncing between Eastern Europe and Middle East and North Carolina and keeps hopping around, we start worrying more. So you can do some things around IP ranges, which takes that friction away from the user altogether, and so those are some of the things that I would encourage admins to look into. Now, sometimes it doesn't work because of mobile providers or other network infrastructure, but it's also potentially a low-calorie way of really sort of avoiding at least the step-up part. They'll still have the MFA requirements. Jay Hurst: And one thing I'll add to that. When we add these types of controls to ... admins are human like the rest of us and we always want to make it as frictionless and easy as possible, but we don't want to trade off the security for it. So IP ranges specifically have been a little bit of our bane for the last six months, because we have a lot of customers who have said, "Well, I will just put in the entire internet range of IPs and go 000 to 255, 255, 255, 255. And then no matter where I'm coming from, I'm not going to get challenged and it's going to be great." And that works, or worked, but it's not secure. And so we don't want to trade off security for just checking the boxes. There are reasons for us doing this. Now, with that one specifically, we have put in changes in place where we won't let you use such a broad internet range to get around it in that respect, where you actually have to put though into like, "Well, what is the internet range that I'm actually coming from and using?" And we have some friction still with that, I would say, with admins using very global deployments or salespeople that travel around a lot and using different VPNs. So we're working through that as one of those compensating controls rather than the primary control. But again, it is up to all of our admins to really think through. We don't want you to just do the check the boxes that we're telling you to check because we want you to check a box. We really want you to think about these changes and what it means to your organization, why we're actually trying to implement these, and ultimately determine what makes the most sense for you and your company. And include your security officers and your CISOs to really validate what meets the requirements that your company has, so that everybody is protected both from a security perspective and ultimately a legal perspective as well. Mike: Yeah. No, absolutely. I mean, we did a security day at TDX, our team did, and we had a CISO in there and a Salesforce admin as one of the kind of breakout topics, and they were in lockstep with each other, and I've always said that. I remember the instance I was managing, once it grew beyond 10 users, I was like, "I really need to figure out what our security posture is and who's in charge of this, who sets password complexities," at the time and stuff like that, and it's not adversarial. It's you want to be in alignment with what the rest of the company's doing and also fall in line if there's a SSO requirement. Boy, I was quick to jump on that, because it also made logging in simpler for my users. Jay Hurst: 100%, and that's the other side of this. All of these things do play with SSO. If you have implemented SSO that has phishing-resistant capabilities already built into it, great. We will take advantage of that from the Salesforce side. We'll trust your provider. We'll trust that your security team has already set up what is needed for your company and we'll let you log in. So we don't want to add these additional duplicative type of controls for you, but we also recognize that a lot of our customers, maybe they're smaller customers, maybe they're not the enterprise size of Salesforce and don't have as structured of a security posture. So in those cases,...
/episode/index/show/buttonclickadmin2/id/42105830
info_outline
How MuleSoft Helps Salesforce Admins Build Better Agents
07/09/2026
How MuleSoft Helps Salesforce Admins Build Better Agents
Today on the Salesforce Admins Podcast, we talk to Mofeyi Oluwalana, Director of Product Management at Salesforce. Join us as we chat about MuleSoft, Flow, Agentforce, and what happens when agents need to take action beyond Salesforce. You should for the full episode, but here are a few takeaways from our conversation with Mofeyi Oluwalana. MuleSoft creates agent-friendly business processes Agentforce can be truly transformative for our business processes, but sometimes it’s easier said than done—especially when multiple platforms are involved. That’s why I sat down with Mofeyi Oluwalana, Director of Product Management for MuleSoft. As Mofeyi explains, a problem many businesses run into when they’re trying to implement AI is how to enable an agent to engage with something like a process that starts on a payment platform, goes through an OMS, and then ends up with a request to a warehouse for shipping. You can’t just hand them the APIs and expect them to figure out the rest. That’s where MuleSoft comes in. It gives you the building blocks you need to codify your workflows into something an agent can understand. Data mapping for real-time app integration Building an agent-friendly business process begins with data mapping. What do you need and where is it located? You also need to understand the triggers that kick off each part of the process. Does it start with a Slack message, or when an order is created? Often, these types of business processes don’t start in the Salesforce ecosystem. MuleSoft allows you to translate these business requirements into APIs that an agent can use to take action. How to present to stakeholders As a product manager, Mofeyi frequently gives presentations to stakeholders, so I wanted to know if she had any advice for admins. “The most important thing when I walk into any room, regardless of the stakeholder, is who are they and what do they care about? Your ability to persuade and influence is largely due to your understanding of the three things that the people that you're talking to care about and how you align what you're talking about with those three things,” she says. Make sure to listen to my full conversation with Mofeyi for more about MuleSoft and mapping your business processes. And don’t forget to subscribe to the Salesforce Admins Podcast for a new episode every Thursday. Podcast swag Learn more Salesforce Admins Blog Post: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, we're joined by Mofeyi Oluwalana, product management director at Salesforce to talk about MuleSoft, Flow, Agentforce, and what happens when agents need to take action beyond Salesforce. Now, for Salesforce admins, this conversation matters because agents are only as useful as the data, the actions, the permissions, and the business logic they can safely reach. So Mofeyi's going to explain to us why integrations are not just about moving data from one system to another. They're about helping Salesforce connect to the real processes your business depends on. We'll talk about the questions that admins should ask when working with other platform owners, how to think about triggers and data mapping and why business context is so important when you're designing actions for humans and agents. So give this episode a listen, subscribe wherever you get your podcasts, share it with another Salesforce admin who's maybe thinking about how Salesforce could connect to the rest of their business. And with that, let's get Mofeyi on the podcast. So Mofeyi, welcome to the podcast. Mofeyi Oluwalana: Thank you so much. Mike: I'm excited. We don't do a whole lot of episodes outside a core platform. And I know my Salesforce admins ... My ... The Salesforce admins of the world work everywhere and especially now with Agentforce and agents and really just the whole bringing people together and giving a complete view of data, integrating data and working with MuleSoft and tools like that are super important. So I'm glad to have you on. I'm glad we got connected, but let's start off and learn a little bit about you. What was your path to Salesforce and becoming a product management director? Mofeyi Oluwalana: Yes. So I started at Salesforce a little over four years ago. I actually came through the APM program, also known as the Associate Product Manager Program. It's a two-year rotational program for folks coming out of university, getting their start in product. So I was able to come in through that program. I did three rotations all across Salesforce. I spent some time in Commerce Cloud working on how to run promotions on Black Friday. Mike: Oh, holy cow. Mofeyi Oluwalana: And I spent a lot of time in experience services, thinking about how we can make it easier for marketers to build nice looking emails. And I ultimately landed in MuleSoft, which is where I continued working with the MuleSoft team on how to make it easier for our customers to connect the systems that they need to, both for traditional use cases, but also for the agentic ones as well. Mike: Yeah. It's funny, for the longest time in my career, I worked retail and the day after Thanksgiving, Black Friday was always the busiest day. And then I got into tech and I was like, "I don't have to work Black Friday." Mofeyi Oluwalana: Guess what? [inaudible 00:03:17]. Mike: And then realized that like, "Oh man, I was a retail associate. There's still computers that have to be had." So I have a soft spot in my heart for all of the people that on Black Friday have to sweat it out and make sure that servers don't go down so that checks and barcodes and everything can happen on the other end. Mofeyi Oluwalana: Yeah. I never realized how complex it could mean to calculate hundreds of thousands of carts when there were promotions running on Black Friday, but certainly it's more complicated than one may think. Mike: Right. And then there's people like me that add stuff to a cart and I'll come back to it a little bit later like, no, there's somebody ... You're using up the last bit of ... I don't know. He probably doesn't know that. And then I'm also old school retail when you used to have to verify checks and put them through little check readers and now everything's a credit card or a tap or thumbprint. Mofeyi Oluwalana: All of the above. Mike: All of the above. But it's all of the above because we're integrating data and we're doing stuff with MuleSoft. So I threw together some questions, but let's start off with where should admin start when thinking about products like Flow and MuleSoft and Agentforce together? Mofeyi Oluwalana: Yeah. The journey with MuleSoft starts the moment that your agent wants to do anything outside of Salesforce. I think MuleSoft is the connector, so to speak. If you have data that your agents need to reach in your ERP, if you have actions your agents need to be able to take with your payment provider, how exactly will they be able to do that? Now when agents and agentic technology became a thing, a lot of people were of this opinion that, well, agents are smart enough to just understand platform APIs or where agents can read the specifications, figure out the commands that they need to execute and do all the things that they need to do to be successful. We very quickly realized that agents are really smart, but at the end of the day, they're limited by the information that they know and the tools that they have access to. So it's not sufficient to just give your agents a platform API spec. That's not going to cut it. Your agents need integrations and APIs that they can directly invoke that map to your business logic. They need this because not every organization operates the same. For one retail provider, creating an order is very different from other retail providers. And understanding that business vocabulary is really important when building agents that actually bring value. And that's really where MuleSoft and Flow come into the picture. You have actions that your agency to be able to take in order to execute a certain business process. Well, that action is really just an API and we serve those APIs so your agents are able to do those things. Mike: Yeah. I've always thought of Josh who's on my team keeps us in check with helping us understand stuff. And what you just described would be like as if you brought somebody new on and said, "Well, here's a phone and a phone book. And if you have problems, the phone number's in the phone book." And the agent's the same way. If you're just saying, "Here's all the specs and the APIs you can call, go figure it out." It's almost like handing them this giant phone book and just saying, "Well, you're smart. You'll figure it out." Mofeyi Oluwalana: Yeah. Exactly. And the thing is that for a retailer, creating an order may not just be creating that order record. Creating an order may be, oh, I create the order record and then I send a notification out via email that says, "Hey, this is your order." And then after that, I create an invoice. There are a number of steps that happen. And how do you codify those steps? Well, it's not enough to just give an agent, again, the phone book and the numbers. It's much better and much more reliable to give them the actual workflow or the API they can use to execute that entire process. Mike: Yeah. Now, one thing that I've run into, so back when I was doing some integrations as an admin, we had different third party tools that were being used. One thing that I didn't know to ask ... And I'm hoping you can help Salesforce admins like, here's the questions you need to ask. We had an instance and we need to connect it to our financial backend, which ran not on Salesforce. I had to find the owner of that, which I didn't know existed. The other thing I didn't know to ask was I didn't know what to ask. And so as we work through different scenarios, you brought up like, well, they want to take the cart and order the stuff. The orders could be fulfilled in a different system. There could be a warehouse system. What are some of the questions that when the admin's sitting down with a business owner and they're saying, "Here's what I want in Salesforce and I'd really like to be able to pull up the orders." And the admin's like, "Okay, now I need MuleSoft because I need to integrate another system and I got to go talk to that systems' owner." What are some of the questions they should ask and what is some of the information they need to provide so that they can start setting up that relationship? Mofeyi Oluwalana: Yeah. I think it first starts with defining what data are you actually looking for? A large part of building integrations between these two systems is I have some data in the financial system. I have my CRM data in Salesforce. How do those two things map together? So I think the first step is what data are you looking for? And what is the mapping between what you're trying to do from a CRM perspective with whatever may be happening in that financial system? So to me, it starts with the data. I think the next question to ask is what is the trigger? It starts off this process. Is it someone sending a message in a Slack channel? Is it this order being created within your ERP? What is it that triggers this process to happen? You build the trigger, then you identify what the data mapping is between the source being that financial system and the target. And that's really how you get towards that working integration. Mike: Now, when I sit down with another platform owner and I talk about using MuleSoft, we can bring ... And so this is where you're going to quickly hear Mike's knowledge run out of gas because here's where Mike's knowledge hits the wall. But we can bring that data in and it can be like a pane of glass. We can just view that data or we could actually copy that data over so that we can use it in reports. And then we could also extend that conversation with the platform owner and say, "Well, if somebody wants to change something, they could change that in Salesforce and then Salesforce with MuleSoft can push that data change back to that platform." How'd I do? Mofeyi Oluwalana: I think you're almost there. Mike: Okay. Fill me in because what I don't know is what I don't know and I feel like there's other admins that may not know this as well. Mofeyi Oluwalana: I think the best way to think about MuleSoft is that real time app integration layer. So not necessarily the data that you want to live in Salesforce. Getting that unified customer view is really important. And so you'll need the data connectors in place to get data from external systems into Salesforce. That is one use case and one that Data Cloud solves really well. The problem that MuleSoft solves really well in tandem with Agentforce is the agent needs to take real time action. Again, so let's say we're going back to the retail example where I create an order. If order management doesn't happen in Salesforce, creating an order doesn't necessarily mean starting that process within the Salesforce ecosystem. I just want to be able to kick off that process with my OMS, my warehouse, my shipping provider. Then in that case, it's real time app integration. It's using the APIs that MuleSoft has to be able to kick off that workflow and that process using your agents. So it's a little bit different than moving data just from one place to another. Mike: No. I think that's also very important because companies have spent money on those systems. And I remember when I was sitting down with our finance team, they're like, "Well, if this data lives here and in Salesforce, what's the source of truth?" And we set it up so that we were just viewing the data because we really wanted sources of truth within the organization. But sometimes the user lives and dies within one interface and where that data comes from, we don't need that exposed to them. Mofeyi Oluwalana: Exactly. And I think it's the use case for consolidating data into one place is a little bit different from the use case of making sure that your agents can take action. And so I think as part of that rationalization that admins and IT teams have to do is what data do I need to drive what? And what actions do my agents need in order to be successful? And this part of that then can separate it into those two camps and then move forward with the tools that help them do so. Mike: Cool. Now, when you connect MuleSoft with other systems, does the admin need to ask for, can you provide a certain API? That's how it's set up, right? Or walk me through it because I've never set it up and I feel like I'm completely happy being the host of the podcast that asks the questions that admins are like, "Oh, I'm so glad he asked that because I didn't know to ask that question and I didn't want to look stupid in front of my IT people." Mofeyi Oluwalana: Yeah. So I think from the business side, we set the requirements about, okay, this is what I need. This is the process that I need done. These are the steps that should happen. This is what should trigger that. You do that definition. Then you can work with the IT team or the API developer to actually make that happen. Now the API developer has access to a number of connectors within MuleSoft that enable them to build these APIs and integrations without relying on what is the API specification for my OMS system or what is the API specification for my [inaudible 00:15:07] system? There are connectors that have a ton of different actions that you can orchestrate within a MuleSoft integration to get a certain job done. So one example that I like to use is the Salesforce connector. The Salesforce connector has a ton of different triggers. Let's say when a new object is created or on when a record is modified that can trigger the integration. The Salesforce connector also has a bunch of actions or operations. So I can create a record in Salesforce. I can delete a record in Salesforce. I can create a topic in Salesforce. All of the different actions that you can execute through Salesforce APIs extrapolate it into a connector so that it's easier to build and then manage these integrations. So the API developer is in charge of translating those business requirements into the integration composed of those connectors that allow you to achieve exactly that. Mike: Gotcha. Boy, there's a lot. It feels like there's a lot to it and we make it look so simple in demos. Mofeyi Oluwalana: Yeah. I will say it's much easier looking at a screen and hearing an audio podcast of what this really is. But I like to think of them as MuleSoft gives the tools, those building blocks. You can compose those building blocks in a certain order to create a workflow. Those are the integrations that you can use and your agents can also use to take action. Mike: Yeah. I always think of it as admins are managing users and agents and agents are just like users. They have to have instructions and guardrails and guidelines for how to operate and what they can and can't do same way that people have as well. I'd love to know, so as a product manager, you got into the program. How has being a product manager helped you understand the product, but also given you a different perspective of working through explaining different customer stories or use cases? Mofeyi Oluwalana: I would say the best, and I'm obviously biased, but the best part about being a product manager is just the sheer amount of data points that you see. One of the things that I think is really important as part of being a product manager is meeting with as many customers as you possibly can because it's through those customer conversations that they're very clear themes that you can draw, not only of how people use the product, but what the product doesn't quite solve for yet. And so going through those conversations has helped me not only understand those customer pain points better, but then understand the things that we need to do in the product to make it even easier for our customers to use. Mike: Yeah. No. I often feel when I was exposed for a brief six months as a consultant that a lot of the pain points customers have always felt the same. I was like, "Man, they have the same problem that XYZ has." And at the end of the day, you'd look at all the customers you manage and you're like, "They have these five problems." And then brand new somebody would come in and you're like, "And you have completely different use case and requirements and they don't fall into those same five problems that I've been solving all day, which is nice." So I'd be curious, as admins are looking more and more at building ... They build agents and apps on the platform and they extend the platform and they use MuleSoft. What are things that you're thinking about as a product manager for, wow, I have to be a year out or two years out. How should admins be thinking about their building and management of the Salesforce platform as a product? Mofeyi Oluwalana: Yeah. I think the most important thing is not just building for where we are today, but building for what's coming. I think one thing that agents in AI technology has taught us is that the technology market is changing so fast. It's going to be a new model. It's going to be some new capability that you want to be able to leverage. And in order to do so, the foundation and the platform that you're building on has to be interoperable. It has to be able to withstand some of those changes that you want to make. The example I always like to give is if having to change a model is going to disrupt your operations, that is an issue. And we know that is an issue because LLM models, a new one comes up every day, another one is deprecated every other week. So I think the most important thing thinking about the platform is how do I make the platform resilient and flexible so that as these new...
/episode/index/show/buttonclickadmin2/id/42018935
info_outline
How Can Salesforce Admins Find Simple Agentforce Use Cases?
07/02/2026
How Can Salesforce Admins Find Simple Agentforce Use Cases?
Today on the Salesforce Admins Podcast, we talk to Kacie Molina, Salesforce Consultant at Kawaii Cloud. Join us as we chat about how admins can start small with Agentforce and still make a big impact. You should for the full episode, but here are a few takeaways from our conversation with Kacie Molina. Why you should start small with Agentforce With all the capabilities Agentforce brings to Salesforce, it’s easy to dream big. Big projects, however, require major organizational investments in time, planning, and execution. My guest this week, Kacie Molina, has some simple advice for her clients who want to get started with Agentforce: start small. Go back to the basics of listening for pain points and building solutions. Once you start piling up small wins and meaningful changes, it’ll be easier to get organizational buy-in for something big and bold. Admins should listen for problems users already have Agentforce solutions don’t always have to be some sort of major business process overhaul. Those types of changes require layers of approvals, budgeting, and business analysis. You end up spending as much time rethinking the business process as you do worrying about executing everything in Salesforce. Instead, look for simple use cases. For example, implementing an agent summary field to help users easily see what’s going on with an account without having to scroll through a bunch of records. “All of those little use cases that maybe are too specific for a flow,” Kacie says, “we can solve them with Agentforce and natural language.” And over time, the simple solutions add up. Keeping AI work inside Salesforce supports better security Even if you’re not already using Agentforce, people in your organization are already using AI. The problem is that they’re often exporting data over to their LLM of choice, which creates all sorts of security vulnerabilities. That’s why Kacie recommends enabling an agent to assist your users. You can use your model of choice, and it’ll be grounded in your Salesforce data and protected by your security configuration. “It shows users that there's a safe place to get the answers they want without having to worry about breaking company policies,” Kacie says. Make sure to listen to my full conversation with Kacie about how to start small with Agentforce. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce 360 Blog Post: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, we're talking with Kacie Molina about how admins can start small with Agentforce and still make a big impact. Kacie shares why the best AI ideas often come from listening closely to everyday user friction. The scattered preferences, repeated questions, messy exports, and those tiny gaps that slow teams down. We'll talk about building agents that are grounded in Salesforce powered by Flow and designed with trust and security in mind. Because the next generation of admins isn't just adding features, they're orchestrating safer, smarter systems where humans and agents work together. So give it a listen, click that subscribe button, share this episode with an admin who is ready to make AI even more practical for their users. And let's get Kacie on the podcast. So Kacie, welcome to the podcast. Kacie Molina: Thanks, Mike. I am super excited to be here today. Mike: Well, I'm excited to get you on. I was a rare occasion scrolling through LinkedIn and I came across a post that you had put up and it mentioned the AgentforceNow workshop that I think you were in one of mine or you'd been to one. Kacie Molina: I was in one of yours and it was done extremely well, might I say. Mike: You were. Oh, good. Well, I try hard. And you're talking about coming up with ideas and you'd had some ideas for agents and I thought, "Man, I just think we need some sort of agent brainstorm podcast." And so that's why I wanted to get you on. But we're going to talk about that. But before we get started, tell me a little bit about yourself, like how you got into Salesforce, what you do, all of the fun stuff. Kacie Molina: Absolutely. So I grew up in more rural USA, an Amish type community, so horse and buggy. Think maybe Little House on the Prairie, early 1800s. So did a lot of horse training and cheese making with my sister, tended gardens and animals. So my skills when I became an adult were fairly traditional. So as I was trying to translate, "What do I use my skills for?" I started nannying, childcare, and then I started my own sewing business. And during that phase, I was actually so bad at running a business, so I knew how to sew, but I didn't really know the business part. So keeping track of contacts and invoicing people on time was such a disaster for me. It was all little handwritten notes. So I was listening to podcasts and I heard about Salesforce. I logged into Trailhead and fell in love with the gamified learning. So it was kind of no turning back for me. I got my admin certification in about three months, landed a junior admin role shortly after. And it's been about five years and 10 certifications since then of working on all kinds of projects from in-house to consulting. And I just think like Flows, all think Salesforce. It's so much fun. And in the day and age of Agentforce, even more fun nowadays. So that's kind of how I stumbled onto the Salesforce path and it has been way too much fun. Mike: Oh man, I bet we could do a whole podcast on the skills of sewing and how they carry over to Salesforce. Kacie Molina: We probably could, although I do think Salesforce might be a little more fun. One, because you don't get poked by needles. That's a huge thing for me. I don't even like going to the doctor and getting poked by a needle and I did that for a profession. So I got poked all the time and it's much less frustrating than trying to deal with broken sewing machines. Salesforce kind of works pretty well. And when you're dealing with old-fashioned sewing machines, not so much. Mike: I could imagine. Yeah. Needleforce, that's what we'll call it. Kacie Molina: Needleforce. Mike: I bet that exists. I think that would be so cool if there was quilting or sewing at Dreamforce. Because I see people on the plane all the time with those knitting needles and stuff. I would love to have the patience to do it. Kacie Molina: It definitely takes a lot of patience. So actually, interestingly enough, when I was just getting into Salesforce, Mallory Donahue, she's such an amazing person I look up to. She was getting into Salesforce at the same time as me and she was also a seamstress. So I think there might be more sewing and needle people out there in the Salesforce world than you might think. Mike: Yep. Well, this is the diversion that Mike's squirrel brain gets on too quickly. But let's talk about ... I was so intrigued by your post, because I feel like even in the workshop that we do, the AgentforceNow workshop that I hosted, I think I'm back sometime in October because I have some vacation in July. We build an agent for Pronto, which is a food service company and it handles questions and cases and order issues. I mean, I feel like by the end of it, it's pretty complex because we've already built some Flows. There's some Apex actions that are in there. And so I feel like we're getting nitty-gritty into it, but your post kind of almost felt like a breadth mint of like, "Hey, have you ever just thought about asking your users five questions or whatever and then turning on Agentforce and giving it on action to do?" Kacie Molina: Yeah. I really like the workshop, because it gets into the more complex things and you see how the broader, the depth Agentforce can go to. But for me, when I started looking at ... Okay. So I serve multiple clients, and a lot of people are excited about Agentforce and I talk to a lot of admins about it as well, but where do you get started? For me, it was the really simple, small things where people I feel like nowadays more and more are expecting to be able to use natural language. And you don't really need to build a crazy, big agent that takes multiple sprints to complete. You can turn one on fairly simply to answer a couple simple questions and people can start to see that magic work from even a small start. Mike: Yeah. And I feel like, boy, this is me just thinking of that post, but you had really good ideas around not just what I would call the hard aspects of business, which is like invoicing and questions and emails, but also kind of like the soft aspects of business, which is just asking Agentforce to coach you that day or just kind of simple questions that you could give it that would maybe make you as a salesperson or customer service person feel better about what you're doing. Kacie Molina: Right. I do. I think it's really cool that you kind of came back to the small ideas and I'm actually very impressed you found that post because I feel like not many people see my LinkedIn post. Mike: You mentioned me in it, so I got that. But I mean, to be fair though, the small little detail, little things like that are actually kind of the bigger hooks to look for. Kacie Molina: They are in there. To me, they're everywhere. So bigger agent ideas, like there's a couple things that can be intimidating about them. One, they take a lot of time. Two, you're always looking for that big idea that is going to wow the whole company potentially. And then it also entails potentially budget and project approval, which you need to get approval for most things that you do anyhow. But the small ideas are something that you don't really need to take a whole ton of business analysis, say, to get started on it. It's something that you could know off the start like, "Oh, I see my users asking about this in a support ticket frequently." So one example recently I had was somebody who said, "We have these client preferences and they're kind of scattered in the system and they're a little hard to find. So is there any way you could summarize them?" And I said, "I think this would be a really good use case for an agent summary field." So you can build the prompt and pull things together. It's the little things like that when you're listening to details that you might not have to go build a Flow or build something so specific for. And that would be my old use cases like Flow all day, every day. But all of those little use cases that maybe are too specific for a Flow to say, "Oh, we'll grab this and explain it this way." It comes out in Agentforce as the natural language and I think that is a huge thing for users that they're looking to see something that feels like them or feels human so they can click onto a record. Say they go to a client account and then they can see at a glance like, "Oh, this person prefers this and this and they don't like this." And that's much easier than scrolling through a bunch of records. Mike: Yeah. I've been so busy building agents that I forgot about some of the other prompt builder and the grounded fields that we had. I used to call them sparkle fields, I think like that, because they have little sparkles next to it when you can automatically have Agentforce kind of fill in based on a prompt and that's exactly it. I mean, I remember rolling out chatter and I did it probably the wrong way and I turned it on for the whole company and the whole company was like 50 users, but I didn't give them any use cases. And if I was now to go back, I would've picked a team of five people and said, "I'm going to turn it on for you and here's a use case to use it for." And I feel like your post was just ... It just made me feel less burdened to come up with these extravagant big ideas to use an agent for and it was more like, "No, you can just have the agent totally answer your front door and be like, 'Hi, we're not home right now,' and that's it." And I was like, "Oh." Because I feel like admins are sometimes sitting down and thinking, "How do I make an agent that answers the questions and closes a $6 million deal?" Kacie Molina: Right. And then there's a bunch of, in my mind, you have more selling at that point. So if you're coming up with a new idea, then you kind of have to pitch it to the company. But something I like about the small ideas I've been picking up on lately is that it's kind of puts us back into our usual admin position where we're hearing problems and then we're offering solutions. So instead of coming up with a new business flow, start to finish of a new agent that can do some flashy thing, which is also extremely cool. You can also take something where somebody says like, "Oh, I export this data and ask ChatGPT about it. Anyhow, I'm already using AI." And so you can start to hear use cases where you're not coming up with something new, you're just solving something they're already doing in a more secure way. And that's another big idea for me is security within Salesforce is so much more attainable for admins. When people start exporting data and loading it up into an LLM that you don't have control over, that's a lot more scary for as an admin than if they can just use it from within the system. Mike: Yeah. I mean, when you and I talked before we pressed record, you even brought that up. If you're an admin and you're in meetings and maybe it's not even a Salesforce meeting and you hear people bring up a different LLM, what in your mind goes off and what do you think of to do as next steps if you were to help coach admins who are in that position? Kacie Molina: Something that's really incredible about Salesforce that maybe not everybody knows that they haven't played around with Agentforce yet is that you can use customized models. So it's not like you only have availability to use the Salesforce model. There's also, you can bring in different like ChatGPT if you like their open AI's model, you can use those different AI plugins within Agentforce. So that's something that goes off my head is if you like something, you don't have to totally forsake that for something new. So you can feel at home within Salesforce with the different AI models. And also, if you're using different AI models, which just like constantly goes off in my brain is that you can do that with way fewer clicks and way fewer security risks if you do it from within Salesforce without having to potentially like fudge on permissions or fudge who has access to it. So that's something that goes off in my head is like, "Oh, if you're doing this with AI already, you can just pull that into Salesforce and then you have even more data." So you are kind of limited when you export data to what the LLM can see. But when you have it within Salesforce, you have so much expansive broader options that you can use. So you can pull in different records and different things from here and there and you don't have to be so specific in one point in time. You can re-ask again in 10 minutes if you have information flowing in. So for me, it's like the different LLMs that are available within Salesforce and data availability is huge. Mike: Yeah. I mean, plus once you export that data and take it out of the platform, for many companies, you're already violating security because that data is housed in Salesforce for a reason and it's not meant to be used elsewhere. Kacie Molina: Yeah, that's a big risk I think with AI. Most companies I think are using it, whether it's an approved tool or not. So when you even enable to me a basic agent in a Salesforce or it shows users, "Hey, there's a safe place where I can get the answers that I want and I don't have to worry about breaking company policies to get the data I need." Because we are in a fast-paced world and I know some users, they really do need to get answers that maybe they can't get within Salesforce today. So they're exporting data and getting their answers with an AI elsewhere. But if they can have that easily accessible and it's safe to use, then I think that enables both admins to feel safer about their org and the users to save clicks, save time and get a better answer Mike: Absolutely. Back to that workshop, I'd love to know your perspective. So we talk about agents and subagents and even in that workshop we create a subagent to handle, I think it's returns and refunds, or order issues and refunds. I'm trying to remember it without remembering it. When you're working with customers or talking with companies about using Agentforce, do you even talk about agents and subagents, or do you just kind of like, "Here's the magic and I'm going to build everything behind the curtain"? Kacie Molina: I usually don't bring up the different agents and subagents and I guess it depends who your stakeholder is. If there's somebody who's more admin-minded and very inquisitive, I would potentially bring it up. But most of the stakeholders, I'm working with bring-me problems and expect me to be the expert and bring them solutions and they don't really care about the very cool details about agents and subagents I think are so fun to talk about. They just want to see the magic on the end. So for me, it's kind of like Flows. I kind of relate most things back to Flows when I'm learning how to code or anything. It all comes back to Flows for me because Flows are magic and they're so much fun. A lot of people don't care if you call sub-Flows or you have different components in the Flow. They just want to know when X happens, they see the screen and then magic happens in the background. And I think agents are kind of the same way where they're really cool to see the backend for people who care, but you can also talk to stakeholders at a higher level without them needing to know that detail. Mike: So you talk a lot about Flows. How important was it for you when Agentforce came out that a lot of the power behind what it could do its actions are powered by Flow? Kacie Molina: Oh, that was huge for me because it feels like it brings the control back to the admin. So I definitely wasn't the most happy, quick adopting of AI when it first came out. I was definitely suspicious and I love my Flows and I love doing things manually. So I was definitely suspicious of AI in general and I've learned to really love it. And part of that is being empowered to do tasks on the backend. So you have the agents that do all of this, we'll say agent magic in the background, but then there's also the pieces that you can build out manually in Flow so you have control over what information you're pulling, what you're feeding to the agent, that way you get the predictive and answers that you want. Mike: Yeah. And even the builder, I feel like that the new agent builder feels like Flow Builder, because, well, there's also variables in there, which still worry me. Kacie Molina: I do like the UI better. Mike: I'm still not the best at better building Flows. I don't know why my Flow confidence is low, but maybe that shouldn't have rhymed, but yeah, my Flow confidence is low. Kacie Molina: I feel like you're probably much more of a Flow expert than you're letting on. I will say though, the new UI gave me a pretty big boost. So the old UI wasn't quite as intuitive or friendly and that's something I really wanted to post on LinkedIn about as people who maybe like me, they got their certifications and started doing their Agentblazer statuses on Salesforce when it was earlier. And then when you come back and you see the new UI and ease of use and having things like variables where you feel more comfortable like a Flow, all of a sudden I was like, "Oh, I can really do this this time. This makes so much more sense." Mike: Yeah. And the fact ... I mean, we don't get into it with the workshop, but I'd love to know, have you ... So the one thing I want to do is I'd love to have an agent help...
/episode/index/show/buttonclickadmin2/id/41920675
info_outline
How Do Admins Set Up Agentforce Coworker in Salesforce
06/25/2026
How Do Admins Set Up Agentforce Coworker in Salesforce
Today on the Salesforce Admins Podcast, we talk to Parth Shah, Director of Product Marketing, and Rikke Hovgaard, Director of Product Management for the Agentforce Coworker team at Salesforce. Join us as we chat about Agentforce Coworker and what it means for the way admins help teams get work done. You should for the full episode, but here are a few takeaways from our conversation with Parth Shah and Rikke Hovgaard. Agentforce Coworker brings AI into everyday business work AI has been a game changer for coding, but integrating it into normal business processes has proved more challenging. That’s where Agentforce Coworker comes in. It’s an autonomous AI teammate, grounded in your enterprise data from day one. For this episode, I sat down with with Rikke Hovgaard, Director of Product Management, and Parth Shah, Director of Product Marketing, to find out what Agentforce Coworker is and how it can help admins bring new tools to their team. Salesforce permissions remain central to trust and access Trust is central to any conversation about AI. New tools are exciting, but you need to be sure that they won’t give people access to data they shouldn’t be able to see. Luckily, Agentforce Coworker works around the permissions you’ve already configured in Salesforce. It’s simple for admins, and keeps those guardrails in place. As far as configuration goes, Rikke and Parth are all about keeping it simple and transparent. While Agentforce Coworker automatically inherits the security and governance settings that are already in your org, you also have special configuration options just for the agent. Admins can connect more data through Data Cloud and Slack For Rikke, Agentforce Coworker is especially useful for summarizing her team’s Slack conversations. She can quickly understand which team member is working on which issue, and where she might need to lend a helping hand. At the end of the day, Agentforce Coworker gives you a conversational interface to do more with your Salesforce data. “Coworker does the work in the background so that your team can actually focus on what matters,” Parth says. Listen to the full episode for more from Parth and Rikke about Agentforce Coworker. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, we're talking with Parth Shah and Rikke Hovgaard about Agentforce Coworker and what it means for the way admins help teams get work done. Now, this isn't just about turning on another AI feature. It's about bringing trusted business context, permissions, data, and automation together so users can ask better questions and take action faster. Parth and Rikke are going to walk us through how Coworker connects to Salesforce data, how it respects existing permissions, and how it gives admins a simpler path to roll out and governance. So, be sure to listen in, press that subscribe button if you haven't already. And I would always appreciate it if you could share this episode with a fellow Salesforce admin. And with that, let's get Parth and Rikke on the podcast. So, Parth and Rikke, welcome to the podcast. Rikke Hovgaard: Thank you for having us. Parth Shah: Thank you. Mike: Well, I'm glad we could get you both on. This is the first time I've had a product manager and a product marketing major on a podcast. Usually, one of them has something too important to do that they bail out, but I'm bringing people together. It's going to be great. So, we're going to talk about Agentforce Coworker, but first I want to learn a little bit about each of you. Parth, I'm just going to start with you because you were the first one to bug me in my Slack DMs about Agentforce Coworker. So, I'd love to learn just a little bit about how you came to Salesforce and what you currently are working on. Parth Shah: Yeah, absolutely. Slack makes it really easy to bug my Coworkers, especially my favorite ones. I sent a lot of emojis. So, sorry, Mike. Mike: Oh, I'm your favorite. Parth Shah: Yeah. So, a bit about me. I grew up in Ohio, was a chemical engineer in undergrad, which really I love to build things and I love to, I guess, solve complex problems. And then when I graduated, I didn't want to go work in a plant, in a chemical plant. And so, I ended up going into sales within oil and gas. Did that for about three years, really understood how to actually talk to customers and what their pain points are. And then switched over to a really early stage startup in education technology, really passionate about education. Did that for about three years, which is where I learned a lot about product marketing. Then got my MBA and then joined Salesforce. So, been at Salesforce as a product marketer for about coming up to seven years. And what's been kind of interesting is I started off on the MuleSoft side of Salesforce, so all about integrations. I really learned what it means to connect data between different systems and how developers and admins go about doing that. And that's when I switched over. MuleSoft wanted to launch a new product called MuleSoft Composer at that time, which is now Flow. So, the next level of after connecting data is like, how do you actually use that data to automate processes and make it super simple for admins like our viewers out there? So, help launch that and then moved over to our Data 360, previously data cloud business. And here, I focus on everything that is around unstructured data. So, again, the through line here, if you look at data is, now we can start tapping into the unstructured data that's all out there because of AI. But the more important thing is now that AI is coming in, how do we actually bring that data, which is full of richness in audio, video, meeting notes, et cetera? How do we bring it into AI so that AI is really grounded? And so, that's been kind of my journey. Obviously, looking back at it, everything kind of connected. I didn't know at that time, but the through line I think is my curiosity of what's next within tech has kind of propelled me to go and find out and work on the cool next product that's coming up, which is also how I land on a Coworker. Mike: Wow. I would've had no idea that you started in oil and gas. Also, what's sales like in oil and gas? Hi, do you want any? Yes. Parth Shah: Cool. Yeah. It's very old. It's a very people oriented business. Mike: Yeah. All right. Well, I had to get that joke in. Rikke, I know you're letting me say your name that way, but in all fairness, being German, there's names that we just can't pronounce. So, I'd love to hear how you got started at Salesforce and obviously you're working on Coworker now. Rikke Hovgaard: Yeah. So, originally I'm from Denmark, which is why my name is a little hard to pronounce. Believe it or not, that's actually a pretty common name in Denmark. So, Rikke is fine here, but I started out after university working for a consultation partner. So, I started implementing Salesforce randomly, started out with Sales Cloud, Service Cloud, moved over to Pardot and Marketing Cloud when that was acquired by Salesforce and ended up also doing a little bit of data and analytics there. And then I decided it was time to move on to other things. So, I joined another partner in London and I moved to London and did that for a bit. I was working a lot on analytics at that point and also marketing. And so, it was mostly focused on what we know now as CRM Analytics and I started a blog on it as well. And so, a few years into it, I joined Salesforce in the CIM Analytics product team, helping customers get the most out of CIM analytics, deescalate some of their issues, best practices, all this kind of stuff. And I think that's where I also got a lot more involved with the community. And then I joined later in Data Cloud, moved over there and about two years ago I started the team with the search team in Data Cloud and we had this idea of Coworker and I ended up working on the admin side. And today I own the Coworker setup experience as well as search manager. Mike: Oh, wow. I mean, all roads lead through Data 360, literally and figuratively. So, Parth, let's talk about that. I mean, for anybody that hasn't heard, what is Agentforce Coworker and what is it there to help with? Parth Shah: Yeah. So, Agentforce Coworker, it is your autonomous AI teammate and it's available in Salesforce and it's also going to be available in a lot of different services like Microsoft Teams, desktop, mobile, ChatGPT, Claude, and a lot more. Fundamentally, what it is is it's an AI assistant that knows your business from day one. It's already connected to all of your enterprise data, whether it's structured, unstructured, your opportunities, your cases, your accounts, all of that context that you have within Salesforce and then what's connected to Salesforce. And of course, it respects all the permissions when you do connect those systems. So, it's grounded in your context from the first prompt that you ask. Number two is it can also orchestrate a lot of your agents. It can also execute a lot of complex tasks so that way, it moves your work forward. In other ways, it takes action on your behalf and you can actually update your CRM using Coworker. And then lastly, we built that headless. So, what's really cool about this is yes, it's available in Salesforce, but now you'll be able to take that Salesforce experience and that data and the context that you've spent so many years building into wherever you work. Now a question that we usually get is like, why do we build it? What was kind of the reason behind it? What's the pain point that is really trying to solve? And I think if you look at the industry trends, what's been interesting and what's kind of been going on is you've seen AI obviously augment software engineering, right? It's writing code and it's improving their productivity, but when we think about how business users actually do their work, it's quite different than software engineers, right? Of course, AI can help write emails, it can create the docs. We're going into customer service, but it gets a bit more complex because in our day-to-day work, we're switching a lot between different applications. We're copy pasting a lot of different data from one application to the next, for example, from Slack to Google Docs and so on and so forth. And it's not adding as much value as it does for software engineering. And so, we said, "Okay, how do we bring all of that context that you have wherever you work into a product that brings in AI and actually helps you do the work so you can focus on a lot of the other things that are more important, like building your relationship?" And Salesforce being Salesforce, we said, "Let's bring in that context and we have that AI and let's put it into Agentforce Coworker, which is where also Salesforce, our users work." And the second thing that we also did on top is if you think through where most of your work actually starts, which is again, very different than how we think about agents in general, it starts from work. Anytime you have a question about a task, you go to search. And so, we said, "Why don't we turn that search into something agentic where search gets you answers and then answers can help you take those actions?" And so, fundamentally what we've done is we brought in that context and made it really easy for our users to use Salesforce and fundamentally get a lot more value out of Salesforce and all the things that it can do with the data that you have. Mike: Yeah. I think what's cool is Coworker's not just for admins. I mean, we'll talk about that, but it's for all of your users. So, one question based on everything that you said, just to make sure that I kind of wrap my head around it as an admin, if I'm setting it up, I can permission what it can and cannot have access to in terms of data. So, if I have, let's say payroll data in Salesforce, I don't want to give it access to that so that Bob and sales can go down and be like, "Hey, I wonder what Sally makes," and then ask Coworker and get a response, right? Rikke Hovgaard: Well, we're actually using the same permissions that you've already as an admin set up in Salesforce. Mike: Oh, great. Rikke Hovgaard: So, there's nothing extra to set up. We tried to make the setup experience super simple and one of that is also, of course, knowing what you've already set up in terms of permissioning. So, if you have in your object given access to something, then the user that is locking in will be able to see that. And if they don't have access to it, they won't be able to see that data either. Mike: Gotcha. Wow, that's huge. So, it automatically just respects the user's permissions as opposed to me the admin having to go and set something up for it. Rikke Hovgaard: Correct. Yeah. Mike: Wow, that's really ... I mean, what's funny is every time I see or work through some of these new products, I'm like, "Oh, well that totally makes sense." Except I think that was probably really hard to do. So, it's always really cool when it happens. It's like intermittent wipers, I suppose. Rikke Hovgaard: Yeah. Well, we have some amazing engineers that figure out all the architecture and build this stuff. So, I can't take credit for that, but I guess from my point of view at least, as an admin, there are so many things that you have to learn with all these new features. And so, our vision was really to try and make the whole setup simple. We also wanted to be transparent around the permissioning and the governance around it. So, we've also tried to balance that, but it's also how much information do you give upfront? How much is in the documentation? So, it's a fine balance to try and make it all simple enough. Mike: Yeah. Now can we talk a little bit about the design of it too because I've seen, thankfully I've had an org and I got to play around with it and it couldn't be more intuitive. I'm sure it could be, but can you talk about what went into how you thought about what that user experience is looking like? Rikke Hovgaard: So, I think it's really down to the jobs to be done. What are you trying to achieve? And that's why the ask mode is very much focused on just asking the question and having that conversation. We wanted to make that a very clear, crisp understanding of what is it that you're supposed to do on this screen. I don't know if you had a chance to also play with the other part, which is we also introduced a new search experience and here that's more of a balance where we're trying to figure out, well, all the things that people love from the original search and how do we bring that in as well as also how do we make it a little bit more new and look a little bit more or nicer, so to speak. I think that has been a little bit of a challenge when we're thinking of all of the data that you can bring in Coworker compared to what we have in Global Search. And now I know we're going a little bit more into the admin side of things, but in order to display things correctly, we need to make sure we also have the information of the data in order to show it correctly. So, that has been an interesting challenge to look at, but yeah, really trying to make it super easy for users to know what they actually have to do, prompt them to get the most out of the tool. Mike: Yeah. I mean, it almost feels like at times there's a balance between making it super simple and having people trust it and making it a little bit more difficult just so that people trust it because you've gone through it. And if it's almost too easy, then it's like, "Hmm, but did it really do? Is it really..." I can always feel that anxiety in a user or an executive when you demo something the first time and they're like, "Yeah, but does it really do that?" Rikke Hovgaard: Especially also when we think about AI, we're putting a lot of trust into it and allowing us to get the answers that we're looking for. And so, I think that's why the whole thinking steps that we have implemented is super important for the agentic response. So, we built a level of trust or transparency in what is it that we're actually looking and giving that information to the end user so they can see how we got to the result. Mike: Yeah, absolutely. Now trust and governance is top of mind. It's literally the first thing everybody talks about with AI. Can we talk a little bit about how that works with Agentforce Coworker and maybe some of what you're seeing in the beta version? I feel like I was on that Agentforce Coworker page and there was an admin on that page that's awfully familiar to a lot of the community that had some great quotes. Rikke Hovgaard: Mike, let's maybe start with how do you actually set up Coworker. I think that will lead into how do we control the permissions and- Mike: Yeah, let's start there. Rikke Hovgaard: Okay. So, the vision we had with the setup page, because if we're being honest, we are using features from other parts of Salesforce. So, we're using Agentforce, we are also using Data Cloud, we're also using global search. So, the thought that we had in the beginning was that's a lot of steps and hoops for an admin to go to in order to set up a Coworker. And so, our vision was really, how can we simplify this? How can we get an admin up and running super-fast? That doesn't mean that they don't have more things they can do, but from the first click turn on to they have access to Coworker, we wanted that to be super simple. So, we have defined a bunch of steps that has to be done in order for you to use Coworker. One, for instance, we have to create the search agent. So, when you click turn on, you actually go ahead and we've automated a bunch of steps for you that goes ahead and creates the things that we need to get Coworker up and running. And in that we've created the agent, we created the permission set group that you need to give access to users and we also have enabled your Salesforce data out of the box. So, any Salesforce data, any custom object, any default or standard object that you have is automatically searchable just by hitting turn on. Now, since we are in a beta, we still need you to opt into the beta. So, there is a little step that some people might miss, which is just go ahead and opt in so you have the new search experience in the front end with the global search bar and the ask button. And then the final step in order to get a user to access and use Coworker is to assign them this permission set group. Essentially, if you're an admin, you're probably familiar with different permission sets are giving you access to different things. So, you need to have a permission set for your agent, you need to have a permission set to use some basic features of Coworker and there's some other things, but essentially we're bundling all of this together in the permission set group. And so, all you have to do is directly from the page, the setup page. You can assign your users directly from that, or you can use the permission sets or permission set group from the user setup page as well. So, that's to say that we tried our best to make everything as simple as possible and then you can lay additional features on top of that with external data, et cetera. So, this is how we are controlling whether or not you have access to it, just permission sets and permission set groups. Mike: That's very easy. So, if an admin turns it on, it's not just available to everybody. Rikke Hovgaard: No. Mike: You can choose whether or not maybe you want a small set of users to try it out first. Rikke Hovgaard: Exactly. Mike: Perfect. Rikke Hovgaard: This is all permission based. Mike: I like it. Parth Shah: Now Mike, following on that, I think if we look...
/episode/index/show/buttonclickadmin2/id/41794575
info_outline
What are Security Essentials for Salesforce Admins
06/18/2026
What are Security Essentials for Salesforce Admins
Today on the Salesforce Admins Podcast, we talk to Laura Pelkey, Director of Customer Security Communications and Engagement, and Sabrina Simeroth, Product Manager for Security Center at Salesforce. Join us as we chat about security essentials for the summer and how Salesforce is helping admins protect their data. You should for the full episode, but here are a few takeaways from our conversation with Laura Pelkey and Sabrina Simeroth. Why admins are a target Security doesn’t take a summer vacation. And while AI is helping all of us do more things faster than ever before, it’s also helping hackers deliver new types of targeted attacks at scale. And as privileged users, admins are finding themselves in the crosshairs. Luckily, Salesforce is rolling out key security enhancements over the summer to help you protect your org. I sat down with Laura Pelkey and Sabrina Simeroth to talk about what threats are out there and how you can be prepared. MFA requirements to protect your credentials Because AI makes it easier than ever to imitate someone’s writing style or even their voice, the biggest threats that Laura and her team are seeing are different variations of phishing attacks. In the end, it’s all about getting someone’s credentials and then using that access to do damage. That’s why Salesforce is requiring all customers to use multi-factor authentication (MFA), as opposed to the gentle nudging we’ve done in the past. However, MFA can still be vulnerable to man-in-the-middle attacks, so admins and other privileged users will need to use a stronger phishing resistant MFA. Finally, Salesforce will require step-up authentication for users attempting a sensitive or unusual action, like exporting a large file. I don’t have to tell you that these kinds of changes can often be met with resistance. Laura recommends framing things in terms of what they protect your users from. Does a salesperson really want a hacker to email everyone on their contact list from their account? It’s not about making you jump through hoops—it’s about protecting you from real risks. Security Center Essentials and Health Check give admins a central view We also checked in with Sabrina about how her team is trying to make it easier to get a handle on essential security configurations across the platform and what settings are most critical. Coming in July, the new Security Center Essentials will let you see everything in one place instead of having to wade through a bunch of permissions and toggles. Health Check will help you prioritize which changes will get you the biggest bang for your buck, and help you track your security posture over time. “It’s all about allowing admins to navigate the security space in a way that helps to reduce the complexity and provide some guidance,” Sabrina says. There’s a lot more from Laura and Sabrina about security on Salesforce and what’s coming next, so make sure to listen to the full episode. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce Help: Trailhead: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: This week on the Salesforce Admins Podcast, we're talking security with Laura Pelkey and Sabrina Simeroth. The threat landscape is moving fast and with AI in the mix, attackers are getting better at targeting privileged users like Salesforce Admins. But in this episode, it isn't about fear, it's about readiness, trust, and the systems admins can put in place to protect our orgs. We'll cover MFA enforcement, phishing-resistant MFA, step-up authentication, and how Security Center Essentials gives admins a clearer view of the settings that matter most because today's admin isn't just managing features. They're designing secure, trusted systems that help businesses move forward. So, let's move forward with this podcast and get Laura and Sabrina on. So, Laura and Sabrina, welcome to the podcast. Laura Pelkey: Hello? Sabrina Simeroth: Thank you. Thanks for having us. Mike Gerholdt: Absolutely. Well, it's always good to talk security. I feel like I was watching the news the other day and they talk about kids when they go to school and then they have the summer off and there's that summer slump of information. I feel like maybe over the summer we kind of have a security slump because we're taking time off and we're going to water slides and watching tornadoes in the Midwest, but maybe I'm just crazy thinking about that. Laura, why don't you catch us up and tell us what's going on with security and some of the newest things in summer 26? Laura Pelkey: I would love to. And I was just going to say our catchphrase, which I feel like I repeat every time we do a podcast together and that security never sleeps even in the summer. Mike Gerholdt: Oh, yes. Right. That could also be a fun summer action film. Laura Pelkey: Yes. I would see it. Mike Gerholdt: I would. Laura Pelkey: Yeah. So, I mean, we're halfway through the year, which is crazy already. And I'd say in the last year we have seen the security landscape and the threat landscape shift tremendously. And especially, I mean, I think a lot of people at Salesforce and a lot of people who are listening are very aware of and use on a day-to-day basis, AI nowadays, which is great. It's a very powerful tool. But what we're seeing in the security landscape is that AI-driven cyber threats have emerged and hackers have really up-leveled their ability to create and execute targeted cyber attacks faster than ever before. So, it's kind of crazy the speed at which this is happening. And unfortunately, I think we're going to see their capabilities get better and better as these tools evolve. Mike Gerholdt: Well, that's not fair. We're supposed to use AI for good. Laura Pelkey: Yeah. Well, the good news is, so yes, I agree. The good news is that providers are now using the same AI models or better AI models if you're in the lucky groups to enhance the security of their platforms and products. So, now the playing field has really been evened out. Mike Gerholdt: Well, that's good. Laura Pelkey: Yeah. But we're in a really interesting time right now with cybersecurity and AI. Mike Gerholdt: I mean, I feel like everybody's trying to figure out AI and now that kind of only muddies the water of impersonating other people or the speed at which it can replicate good or bad effective use of, I don't know, text messages or different spoofing. Can you walk us through how does Salesforce leverage AI to proactively block the type of threats that you're seeing today? Laura Pelkey: Yeah. So, there are some really common threats that just across the industry we're seeing that AI is being used for. So, account takeovers and this is a really common one. This is where attacker might use AI to create a phishing campaign, which targets users and tries to get access to their account. And then once they're inside a user account, they can then utilize all of the privileges that that user has to continue carrying out their attack. So, it's really about getting credentials, finding creative ways to get around our existing solutions to get those and then get that user profile and exploit it. Mike Gerholdt: Yeah. I think one thing that you've pointed out in the past is identity-based social engineering. So, I'd love for you to kind of expand on what makes admins and some privileged users such high-value targets for that. Laura Pelkey: Yeah, that's a great question. So, identity-based social engineering is when attackers create very highly personalized phishing campaigns and this can be phishing. Traditionally, we saw that with email five years ago it was primarily email. Now we're seeing phishing campaigns and especially with the help of AI, attackers are conducting these campaigns over voice calls. So that's called vishing or voice phishing. And admins are such a appealing target for these attackers because they're privileged users, they have such a high level of privilege and they can do almost anything within their Salesforce org. And so, that persona in general is highly targeted by attackers because of their level of access. Mike Gerholdt: What would you suggest an admin listening to this do today to start just being ready for if that phone were to ring and it was a malicious actor on the other end? Laura Pelkey: Well, I mean, we always say use common sense. So, there are a lot of built-in protections that I can talk about within the Salesforce platform that will help protect admins, which is great and I'll talk about those in a minute. But if that ever happens, if you are an admin and you get a call from a number that you don't know, from a person that you don't know claiming to be your help desk or claiming to be even Salesforce in some cases and they're asking you to log into your Salesforce account for them so that they can help you do something, that's typically how these things are positioned, that kind of scenario is never going to legitimately happen. No one's going to call you or email you and send you a link and ask you to log into your Salesforce account ever. That's just not something that we do. So, look for red flags like that. And in general, whenever you're suspicious about something from a cybersecurity perspective, the best thing to do is disengage with whoever is trying to speak with you or engage with you and get you to do something and then hang up the phone, don't email back whatever it is and then call your help desk line or call your Salesforce account person in a number that you already have and that you are already confident in and that way you can verify if that interaction was real or not. Mike Gerholdt: Yeah. I think the thing that always struck me when you would bring it up or somebody from your team, Laura, would bring it up is it's that human nature side that they try to take advantage of like, "Oh, you want to be helpful and you don't want to cause a problem." And so, you just kind of go along with it and it's like, yeah, you're not causing a problem if you just say, "Let me call back the number that I trust and talk to somebody there." Laura Pelkey: Yeah. Yeah. And you can do that in any scenario. If it's like your bank, someone calls and says they're from your bank and they're reaching out to you, that probably isn't going to happen. So, what you said, Mike, is exactly right. The reason that people are exploited is because we have these human behaviors, which human behaviors are great. They make us who we are, but false urgency is one of these tactics that attackers will use and they'll try to get you to do things quickly without really thinking about it. So, if we feel like goosebumps or we kind of have a weird feeling about why is someone calling me and asking me to log into my account, we can just stop and think and that's the best thing you can do to break this cycle. Mike Gerholdt: Yeah, absolutely. Well, let's talk about some of the security enhancements that has come out over the last few years. I mean, we have MFA and we have a whole bunch of other things. Laura Pelkey: Yeah. Yeah. So, we're very aware that this threat landscape has shifted and that it's our duty to protect our customer's data as it always has been at Salesforce. We take that super seriously, trust. Trust is our number one value. And so, to better help our customers protect themselves against these quickly evolving threats, we have some new enhancements and some new requirements in our platform that will help harden our customer security posture against these kinds of attacks. So, one thing, I'll only talk about a few, but one thing, and everyone really should be using this because it's been around for a while, is MFA, like you said, Mike. And I think we had started requiring customers to use MFA back in early 2022, I think. So, it's been quite a while, but we're actually going to be enforcing that. So, rather than being contractually required, it's going to be enforced in the product, meaning you may not be able to log in if you are not using MFA. So, this is going to happen starting in June. There's a phase rollout based on sandboxes and production environments. So definitely check out the documentation for the exact dates, but it is starting now. So, soon all Salesforce customers will need to use MFA to log in. Mike Gerholdt: And I know we've done security workshops. I've seen you present a lot. One of the questions, and inevitably the person that is listening to this is like, "Yes, Mike's going to ask this question." That comes up in all of the security workshops is, "Well, we use SSO to sign in. Do I still need to use MFA?" And we just threw out a whole lot of acronyms there, so I'll have you explain them too. Laura Pelkey: Yes. So, SSO- Mike Gerholdt: Alphabet soup. Laura Pelkey: SSO is single sign-on. That's a great question, Mike. So, if you do use SSO to access your Salesforce instance, then that does satisfy the requirement, but we ask that you also use MFA on top of your SSO. SSO is great. It's a great tool to make it possible so that you don't have to remember a million passwords for all of your work apps. It's awesome. Just have to remember one, but as long as you are using MFA to access your SSO, you are satisfying this requirement as well. Mike Gerholdt: Awesome. You explained it and you used all the alphabets too. Laura Pelkey: All the letters in the alphabet. Yeah. Mike Gerholdt: It's like MFA is like bacon and SSO is like a cheeseburger. It makes it better. Laura Pelkey: Yeah, exactly. Yes, I love that. Mike Gerholdt: Because MFA is also just good on its own. I mean, I'll just eat some bacon. Laura Pelkey: Yeah. So you can either use MFA for direct logins directly into the UI or with your SSO login. Mike Gerholdt: So, this makes me think of, you talked about phishing. Oh, boy, this is where audio doesn't serve me well. Phishing and vishing, which there's the PH and the V, but because I'm from the Midwest, you can't hear the difference. Could you explain the difference between standard MFA and phishing-resistant MFA? Laura Pelkey: Yes. So, yeah, this is actually kind of a newer... Think of it as a supercharged form of MFA. So, what traditional MFA, you have two layers of verification before you can log into an account. With phishing-resistant MFA, you are actually bound to a specific site when you're logging in, which makes it harder for attackers who are trying to get you to log into those fake websites that might look real. It actually helps prevent that. So, that's why we call it phishing-resistant MFAs because that is a really common thing for phishing attacks to leverage is these man in the middle fake websites where you think you're logging into your actual tool or product, but you're logging into a fake version of it. And another thing that's great about phishing-resistant MFA is most users of this use biometrics as their second form, so like a fingerprint or a retina scan depending on the device that you're using to log in. So, that just makes it, again, harder to fake and harder to intercept. So, this is most companies are moving towards phishing-resistant MFA as it's so effective, and which is why we're actually also starting in June, so this month, we are requiring system administrators and privileged users to log in with phishing-resistant MFA. Mike Gerholdt: And why is that? Laura Pelkey: Because those are privileged users like we talked about, they're highly targeted. They're more targeted than their regular user and they have such a high level of access that if their account, especially were to get compromised like an admin, it could cause a lot more damage. And so, we want those folks to be extra, extra protected. So, phishing-resistant MFA we feel is the best thing to ensure that those accounts and those users stay safe. Mike Gerholdt: I like that. Now I know I've had to roll out different features to users and sometimes you can't do it fast enough because they love it and other times it's getting them to eat their vegetables. For admins that are worried about user friction with MFA, what would your advice be? Laura Pelkey: Well, hopefully your users are already using MFA. If they're not, we often find that if an admin holds a training, some informational sessions with their users, some kind of enablement before this change happens and also talk to them about the value that this provides to them. The last thing a salesperson, for example, would want is if their account got hacked and then their clients are getting emails that are sent from them, which are actually malicious. They don't want that to happen. No one wants that to happen. So, just there's some enablement materials that exist on the help portal that could be helpful, but really it's admins explaining and enabling their users why this is so important and helpful. Mike Gerholdt: One last thing, admins love reports and I feel like isn't Salesforce doing something to help admins set up different kind of reports or report actions to understand maybe when a user would log in that's outside of the normal behavior? Laura Pelkey: Yes. Yeah, I love that. Yeah. So, this is another enhancement that's coming this summer and it's called Step-Up Authentication and we're doing it for report actions and for anomalous report behavior. So, let's say one of your users all of a sudden wants to export a very large file or something. There's something kind of unusual about this. Anomalous is the word we use in the security community. Our platform is now able to flag that and will cause a step-up authentication toggle to happen. And if you're not familiar what step-up authentication is, it's basically like having to re-log in after you've logged in, which I know might sound annoying, but there is a really good reason for this. So, I always think of it as like the airport analogy. So, when you go to an airport, you have to show your boarding pass and your ID to get through the security line, right? Mike Gerholdt: Mm-hmm. Laura Pelkey: And then once you're through security in the airport, you have the freedom to kind of walk around, go get a coffee, go get some food, maybe go to the bathroom and you have freedom of movement because these are not sensitive actions, what we would call in security. But when you get in line at your gate and you're trying to board the plane, you have to again, show your boarding pass in order to get on the plane. And that's actually a step-up authentication. You can think of it like that. This is because getting on the plane is considered a highly sensitive action and requires additional identity verification. And so, that's what we're trying to do within the product. So, say an adversary or an attacker did get access to your Salesforce instance through a user and now they're inside and they're trying to do sensitive things or they're trying to do things that are very bad, we are now able to force them to verify their identity again, which would likely stop that action. Mike Gerholdt: Nice. Well, here's your Starbucks, but you can't board the plane. Laura Pelkey: Yeah, exactly. Mike Gerholdt: I like it. Laura Pelkey: I mean, if you lose your boarding pass in the airport, you can't get on the plane, right? Mike Gerholdt: I mean, you should- Laura Pelkey: Oh, wow. Mike Gerholdt: It also probably means I've lost my phone and the whole day is just going downhill from there. Laura Pelkey: That's a big problem. Yeah. Mike Gerholdt: I know. Laura Pelkey: But that's the thought behind that. Mike Gerholdt: Sabrina, we didn't forget about you. It's usually Laura has a long laundry list of things to talk about, but I'd love to bring you into the conversation. First of all, I think this is your first time on the podcast, so tell us a little bit about what you do at Salesforce. Sabrina Simeroth: Yeah, thanks so much for having me...
/episode/index/show/buttonclickadmin2/id/41682270
info_outline
How Agentforce Helped Build a Food Waste Solution in Days
06/11/2026
How Agentforce Helped Build a Food Waste Solution in Days
Today on the Salesforce Admins Podcast, we talk to Parth Sevak, Director of Technology and Principal Architect at Incepta. Join us as we chat about how Parth built a multi-agent system designed to connect surplus food with the people and organizations that need it the most and won the Agentforce for Good Grand Prize at the TDX Hackathon. You should for the full episode, but here are a few takeaways from our conversation with Parth Sevak. The Hackathon project focused on reducing food waste If you listened to last week’s episode with Alexandra Iyer, you know that Agentforce for Good was a popular problem at this year’s Agentforce Hackathon at TDX. Contestants took on big issues like nonprofit volunteer coordination and disaster relief. That’s why I was so excited to sit down with Parth Sevak, whose project Harvest Bridge won the Agentforce for Good Grand Prize. Harvest Bridge is a multi-agent application that connects food donors with organizations near them. As Parth explains, food waste is a serious problem. According to the UN’s World Food Programme, about 318 million people are facing acute hunger today. “In North America, 30-40% of the food that is produced never gets eaten,” he says. So he decided that this would be the perfect problem to tackle for the Agentforce Hackathon at TDX. Simple integrations and out-of-the-box tools Under the hood, Harvest Bridge features multi-agent coordination between four agents to handle donor intake, food matching, volunteer logistics, and reporting analytics. While it sounds incredibly complicated, Parth is quick to point out that 80% of the work was done in configuration with out-of-the-box admin tools. Parth needed to write some Apex to do specific things like geo-matching, which he vibe-coded with the Claude plugin for Agentforce. Crucially, he didn’t have to write glue code to make everything work between Agentforce, Data 360, automations he built in Salesforce, Slack, and Tableau. “All of it just worked like a charm,” he says, “five years ago, that integration story would have been months, if not years.” How to get started building Agentforce solutions In just a few days, Parth was able to build an autonomous, multi-agent system that uses Agentforce, Data 360, Slack, and Tableau to match surplus food with local organizations and coordinate delivery in under 90 minutes. If you’re looking to get started with Agentforce, Parth recommends jumping on Trailhead as your first step. The Agentforce Specialist certification gives you the tools you need to start building, and then it’s all about getting your hands dirty. Make sure to listen to the full conversation with Parth Sevak about how he built Harvest Bridge and won the Agentforce for Good Grand Prize. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Trailhead: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins Podcast, we're talking with Parth Sevak about how a real world food waste solution went from idea to working system in just days using Agentforce. Drawing from a challenge that impacts millions globally, Parth built HarvestBridge, which is a multi-agent system designed to connect surplus food with the people and organizations that need it most. Now, this was part of the TDX26 Hackathon Challenge and what makes this conversation the most compelling isn't just the technology, it's how admins and architects can really orchestrate data automation, AI agents and human coordination together without months of integration work. We also get into what it means to design a trustworthy system where humans stay in control while AI handles scale and speed. So, if you've ever been wondering about how Agentforce changes the role of the Salesforce admin from builder to system orchestrator, this episode's for you. Be sure to subscribe, share the episode with your team, your friends, your local Salesforce admin user group and let us know what kind of real world problems you'd solve with Agentforce. But for now, we're going to get Parth on the podcast. So, Parth, welcome to the podcast. Parth Sevak: Thanks, Mike. Really glad to be here. Mike: Well, I'm glad to have you. So, if everything shakes out, the episode before this will be the episode with the Agentforce for Good People, but sometimes scheduling is what it is. But I was at TDX this year and we had Hackathon winners and you were part of the Agentforce Hackathon. And so, that's how I got connected with you. But I think before we get into that, I'd love to know a little bit about how you got started in the Salesforce ecosystem, what you do and let's go from there. Parth Sevak: Absolutely, Mike. Again, thanks for having me. So, the story of how I ended up in Salesforce is really interesting. That was not my plan. In fact, back in 2011, I was a fresh computer science engineering grade hunting for my first real tech off somewhere. And that time, the startup economy was just taking off and Java was the thing, the only thing in my perspective as far as I was concerned. And then my boss pulled me aside and said, "Parth, you are going to work on Salesforce." And I said, "Oh, what is Salesforce? I'm looking for Java opportunities." So, he made his pitch. "Apex is basically Java, you will feel right at home. And there is this thing called Dreamforce, massive event, you might get to experience it someday." And that's where you see a lot of enterprise innovation is going to come in. Basically, he sold it and 16 years later, here I am loving the challenge, loving the stretch, always finding ages I haven't touched yet and it keeps me motivated. Mike: He was a heck of a salesman. Java, here's Salesforce. It's just like Java. And if you're good enough, you might get to go to Dreamforce. That's awesome. So, go ahead. Parth Sevak: Yeah. And just to tell you what currently I'm doing. So, these days- Mike: Yeah, please. Parth Sevak: ... I'm working as a director of technology at Incepta Solutions where I lead data, CRM integration and agentic AI transformation for enterprise clients across financial services, retail manufacturing and pharma healthcare. But I make sure to keep one foot in hands-on building always. And that's how HarvestBridge happened at TDX, as you see. Mike: Yeah. So, let's talk about that. I mean, it's really cool. This isn't our first hackathon since we've come out with Agentforce. We've been doing hackathons and the early hackathons, I was a judge in quite a few and some of the really neat ideas that people were coming up with for the use of AI agents and some agentic use cases I think were really kind of interesting. I know we had that at TDX. So, I guess let's start off with the TDX Agentforce for Good Hackathon. What made you want to enter that to begin with? Parth Sevak: Yeah, to be really honest, it was again, my boss encouraged me because even I was not sure whether I would be attending the TDX in first place, but my boss was really motivated to join the party and he invited me. And then I realized, okay, if I register for TDX, there is an option to participate in the hackathon. And that's how my journey began. Mike: Oh, well, we made that awfully easy for you, didn't we? Parth Sevak: Yeah. Mike: I mean, I would like to think I'm a competitive person, but when it comes down to it, I think I really like watching competitions as opposed to competing. What made you think, "Oh, I'm going to get into this because the idea that I have is so great." So, tell me a little bit about what your idea was and what it solved for. Parth Sevak: Yeah. So, honestly, the topic I picked is not a fleshy topic, right? I picked the food waste, which is definitely not a hackathon topic. It's not a biotech, it's not a blockchain. But my heart said, pick it anyway because food is the most basic thing and somehow we managed to throw away nearly half of it while families a few blocks away are skipping dinner. And then again, I am a data guy, so please allow me to share some numbers here. Mike: Oh yeah, please. Parth Sevak: And that application for me, that 2.5 billion tons of food wasted every year, 783 million people are facing hunger. And when I read that and thought, "This cannot be right." And I check UN website, food banks' websites, and again, the numbers held up. And the on that hit really hardest is this number, which is in North America, 40% of food produced never gets eaten right here in our cities, whether it's San Francisco, Toronto, basically with our neighborhood. Mike: I mean, that's crazy. I know I've seen ... I forget the show. It was on Discovery Channel here in the US and they talked about ... I didn't know there were hog farmers out in Las Vegas, but there are and they work with the casinos. The casinos have all those buffets to get all their food scraps. I thought that was really, really fascinating. I guess you think about it and until you look at it on a larger scale, do you even understand how much food is produced and unfortunately how much is wasted? Parth Sevak: Yeah, totally, Mike. Mike: So, let's talk about your solution because it involves Agentforce. So, were you going to feed Agentforce some apples? That's a joke because you can't feed it apples. I mean, you could feed it apples maybe like an emoji. Parth Sevak: So, yeah, I think let me first of all piggyback on this one, this idea that this HarvestBridge as a idea, right? When you look at this project from the outside, it may sound like as a big AI project. We have four agents, multi-agent coordination, the whole team, but the reality on the inside is three of those four agents are about 80% configuration. And this is the Agentforce play in my opinion because I used agent builder topics, instructions, guardrails, then flow for orchestration, custom objects, validation rules, formula fields, like typical standard admin toolkit. And obviously, there was a 20% which is code, which is attacks actions for things like geo-matching. So, that also like I wipe coded with cloud plugin for Agentforce available. And what came back was the quality of a season developer with proper design patents, clean error handling, governance limit aware as I made sure to provide the right context. That is very important. So, anytime if anybody ever built a flow with a decision branch and a few validation rules, you can think the way you need to build an agent. The tools just caught up with you. Mike: Yeah. Well, actually we had a few presentations at TDX and the admin track from admins who are vibe coding and I've had a few people on the podcast talking about how they've vibe coded and just the product itself is just insane to think you can sit down and there's a plan and there's an action button. And even in the plan button, I was really impressed with how much Agentforce Vibes could give you back as an admin to really help you plan your application. Parth Sevak: Yeah, that is really empowering thing has happening just in front of eyes and not just that, but we are at the forefront of that innovation and definitely the Agentforce for Good is really the category that should motivate everybody to get their hand dirty for something to produce something really good. Mike: Yeah. So, let's talk about while you were building your application to deal with food waste, what was a point in time when you sat back and maybe learned something that you didn't know as you were building it? Parth Sevak: Yeah. So, before I answer that question, I need to tell you how I approach the build process if it is okay. Mike: Oh, please. Parth Sevak: Yeah. Okay. So, basically like Mike, to be very frank, I didn't follow a plan. I would love to say that I sat down on day one and map out a clean sprint like user stories and backlog, but that was not what happened. What actually happened is I just started building right away. So, I picked the first thing, which is a donor bot. This is the agent that talks to donors and captures surplus food. First I got that working and the moment it ran, the next question was obvious that, okay, now surplus is in the system in the Salesforce. Now where does it go? So, I built the matchmaker agent and once matchmaker was written in matches, the next question came up on its own that who actually picks the food up and delivers it? Who is the volunteer? That became logistics coordinator on Slack and by then, I had data flowing through the whole system and it was almost asking to be visualized and that become impact analyst on Tableau. So, overall, I ended up building four agents, each one revealed the next. And what made that look possible beyond Agentforce itself is that I had AI assistance alongside me the whole way, like not just for the apex spaces we talked about, but for the thinking too, like should this be one agent or two? What edge cases am I missing? Is this topic structure clean, less a solo build and a more long conversation with a partner who is read every Salesforce document kind of like ever published. So, when I started building on the Agentforce platform, I had some high expectations and let me tell you, like what really surprised me in a pleasant way is the breadth. I use Agent Builder for the agents themselves, the topics, instructions, custom actions. Then underneath, I relied on agent script for the deterministic safety rules. So, things like allergens and expiry windows are not left to the LLM to decide. From there, I brought an Agentforce voice for the call-in donation capture, which I think just became GA recently, which I think, yeah, an intelligent context handled passing food details out of photos that donor send in. All of it was sitting on Data 360, which held the unified profiles of donors, recipients, and volunteers. Here is the part that genuinely caught me off guard. I didn't write any glue code between those layers, like an agent would hand off to a flow which will call an apex action that will update the records in the Data 360 and that would trigger a Slack message. That would feed back into another agent's context and all of it just worked like a charm. And if I reflect like five years ago, that integration story would have been months, if not years and this made me possible to deliver something so amazing in just few days and that was my wow moment from the whole build. And Mike, if I just like tell you another perspective that the other thing I was nervous about going in was hallucination because I am building four different agents and then you have obviously hear the horror stories, right? But between agent builders topic structure, agent script for hard rules and clear instructions, that's how I make sure that my agent behave very well. So, food safety rules around allergens and expiry are not something that I would let an LLM to decide. I would make sure to enforce them deterministically at the architecture level and for that I leverage agent script and thank God, that agent script work really well and where I would love to see things improve is just the debugging because when an agent's reasoning across topics and data sources, tracing what it actually did is still a bit difficult. So, the tooling is getting better, but its place still I'm watching. Mike: Yeah. I do an Agentforce Now workshop every month and I always try to point out when we're building it and preview how you can go through and see what the agent is doing and how it's deploying its sub-agents. And I think with everything we're still learning what AI is and how to debug it and it's this interesting intersection because we want the intelligence and the context and the engagement of AI, but I think people got really hooked on how perfect a bot would respond and a bot would respond because it's always A, B or C and we gave it those answers. And the nice thing about AI is it doesn't necessarily have those answers, but it can think independently, but we still want that control. So, it's like building the perfect employee. Parth Sevak: Yeah, absolutely. And that's why, like as I said, Mike, the deterministic approach is very significant and particularly, I would kind of rely on agent script that really gives that control back to the human or admins or architects so that they don't need to basically do the guesswork that, okay, what will be the response of agent if I do this prompt versus that prompt? It can simply be predictable. So, that is the cool thing I would say about the agent script and what it makes possible now in the reality of governance. Mike: So, I think one thing that's really cool about hackathons is it's kind of blue sky and green field. You can come in with a solution and like in your case, it really solves a real world problem. If somebody is listening to this and thinking, "Wow, I'd really like to get started with Agentforce," what was the first thing that you did to get hands-on with Agentforce and start understanding the product to the level that you do now? Parth Sevak: Oh, I think I don't recall exactly the date and time, but I think it was I think a couple of years ago when there was already like Agentforce discussion has already began because I was active in the community and thanks to Trailhead. And I basically started my journey to equip myself about like what is this best Agentforce, right? And then I did few modules and ultimately I think last December, I got certified in the Agentforce as Agentforce specialist. So, that really gives me clarity that okay, just not about like, okay, what I can build for the demonstration purpose, but also for me, it's very important that I can actually pass on that knowledge to my team so that they can ship the products or ship the agents that is really the kind of like configurable and scalable across the systems, not just one particular cloud or system or CRM. Mike: Yeah. I mean, anytime I need to learn something for a demo or for a presentation I'm putting together, I usually turn to Trailhead first and kind of walk through it and get an idea of what the product can and do and how I can configure it. So, I'm right there with you. Even Salesforce people still learn on Trailhead. Parth Sevak: Yeah, that's amazing actually. I call it as a live report, it's always live. You never get any stale recipe or stale dishes. So, that means everybody, whether it's a employee of Salesforce, partners or customers, everybody can rely on Trailhead for the real time information for sure. Mike: So, moving forward, we have a little event in the fall that we do called Dreamforce that you mentioned. Obviously, none of us have a crystal ball. Well, I'm sure a few people in Salesforce do to know what we're going to show, but what are some of the things you would love to see this year at Dreamforce? Parth Sevak: Okay. So, as I said, I have been part of the Salesforce ecosystem for almost 16 years and what really amazing to see is that the platform never sits still. So, when I started, here is the context, Mike, what I'm talking about is when I started we were building the S controls and it came to Visualforce. Mike: Yeah, that's back when I started too, called them Superman controls. Parth Sevak: Yeah. Then Visualforce, like proper MVC, controllers, pages, then Aura framework. And my first test of real component-based DUI was that Aura and then Lightning Web components. And I remember thinking, okay, this changes everything. And now 2026, we have got React running on Lightning Web Runtime along with the Headless 360 and agents that reason and act on their own and every leap has bigger than the last, right? So, it is the one ecosystem I have worked in that's never stopped becoming more interesting. I will make sure like a kid, I always go without any expectation that, okay, oh, what I'm going to get out of this conference because I know whatever will be there, it will be super amazing, super productive and...
/episode/index/show/buttonclickadmin2/id/41598370
info_outline
Agentforce for Good Shows the Power of Inclusive Innovation
06/04/2026
Agentforce for Good Shows the Power of Inclusive Innovation
Today on the Salesforce Admins Podcast, we talk to Alexandra Iyer, Director of Marketing Strategy and AI Transformation, and Global EVP of Abilityforce at Salesforce. Join us as we chat about the Agentforce for Good Hackathon at TDX and what happens when technology, accessibility, and community come together to solve real-world problems. You should for the full episode, but here are a few takeaways from our conversation with Alexandra Laxmi Iyer. Agentforce for Good expands AI beyond business problems This year, we added a new twist to the types of problems you could solve for at the TDX Hackathon. We called it Agentforce for Good, and it offered participants a way to work on some big problems: food insecurity, nonprofit volunteer coordination, disaster relief, and more. That’s why I’m so excited to bring Alexandra “X” Iyer on the pod. She’s the Global EVP of Abilityforce, Salesforce’s internal business unit for people with disabilities and their allies. They’ve been running an internal accessibility hackathon for years, and Agentforce for Good grew out of a desire to open it up to the community and broaden its scope. “Builders who probably never thought of themselves as social impact developers shipped working Agentforce solutions for all sorts of big problems,” X says, and 62% of this year’s Hackathon entries were solving for Agentforce for Good. Inclusive design creates better technology for everyone “When we are challenged with designing for the edge, you make the center better,” X says. Finding challenges that force you to push tools like Tableau, Slack, and Agentforce in new ways helps you uncover capabilities that you might not have otherwise thought about. As X likes to put it, “Good design is just good design, full stop.” Going forward, X and her team are looking for ways to expand Agentforce for Good beyond TDX. Consider this a call for participation, and she encourages you to get in touch to volunteer or bring Agentforce for Good to a community event near you. Lived experience is a valuable design credential “One of the main things I would love for people to take away from our conversation is that your lived experience is a credential,” X says. “It's not a distraction.” After all, designing around new challenges is how we come up with innovative solutions that nobody else would have thought of. Your unique perspective is valuable. Make sure to listen to the full episode for more insights from X. Next week, we’ll hear from the winner of the TDX Hackathon, so make sure you’re subscribed to the Salesforce Admins Podcast. Podcast swag Learn more Trailblazer Community Group: Admin Trailblazers Group Social Love our podcasts? Subscribe today or ! Full show transcript Mike: This week on the Salesforce Admins podcast, we're talking about what happens when technology, accessibility, and community come together to solve real world problems. I'm joined by X or X, who is the director of marketing, strategy and AI transformation at Salesforce and Vice President of Abilityforce to discuss the Agentforce for Good Hackathon that happened at TDX. Now we're going to find out that it started as an employee-led accessibility initiative. And has grown into a community movement bringing together admins, developers, partners, and customers to build solutions that address challenges like food insecurity, disaster relief, and volunteer coordination. We're going to talk about why designing for inclusion leads to better outcomes, how AI can help scale impact, and why your lived experience might be one of the most valuable design credentials you have. So if you're thinking about how data automation, AI agents work together to create meaningful change not just inside your organization, but in the world around it, I promise you this episode is for you. So let's get Alexandra on the podcast. So Alexandra, welcome to the podcast. Alexandra: Thank you for having me, Mike. Mike: I think it's going to be a fun conversation because we're talking about Agentforce for Good today. So let's get started to learn a little bit about you and your path to Salesforce and your path to the Agentforce for Good Hackathon. Alexandra: Absolutely. Thank you for asking. And by the way, I go by X for short, so feel free to call me X. Mike: Perfect. X: I'm the director of marketing strategy and AI transformation for our customer success and partner marketing team here at Salesforce. And I'm also the vice president for Abilityforce, our employee resource group for people with disabilities and their allies. And when we're talking about the Agentforce for Good Hackathon, this is really a moment where these two worlds actually collided. Normally, they wouldn't collide, but the hackathon is genuinely where they did. I come from 20 years of go to-market experience. I'm not a builder by background. I actually got curious in a marketing role about the products that I was marketing and how they worked. So that curiosity pulled me into the platform, into development, into force certifications that I did not see coming. And then, I went from talking about technology to building on it and that changed everything. I believe that the AI era is only worth building if it's built for everyone. That's not a tagline for me. It's a design requirement. And I truly feel that admins have always known that. Admins are the ones closest to the humans using the platform, the problems, the friction, the gaps. And I also believe they're the closest to the solution. So I think it's not a support function, it's a superpower and I'm really happy to be here. Mike: Yeah. Well, I agree with everything you had to say, X. I do agree that admins are very superpower. I also think just in general, when we think about technology, it's not how do we roll it out for some people? It's how do we roll it out for everyone? And I remember this is a long time ago, but when I first joined a company that did workplace assessments, I remember they said, "Well, when creating the test, we have to assume the person doesn't know how to use a mouse." And I remember thinking to myself, this is ... closed-minded is the wrong word, but I just was unaware. And I was like, "Well, who doesn't know how to use a mouse?" And they're like, "Well, you don't understand." And I was like, "Oh, that is a huge challenge." And sometimes until you get someone else's perspective, do you really understand like, it's not that you've been living under a rock, it's just too often sometimes people assume their world is just like everyone else's world and it's not. So there's a Mike side of I had to remember that not everybody knows how to use a mouse sometimes. X: Well, Mike, frankly, I consider that the main dish. I don't think it's a Mike side. I think that if you're designing for the default user, it's not really good technology and that when you optimize for the average, you're excluding by design. So I think it's the main dish and it's part of the reason why the Agentforce for Good Hackathon even exists. So to give you a little context, three years ago, the Agentforce for Good Hackathon was an internal-only, employee-only hackathon led by Abilityforce. I mentioned earlier Salesforce's business resource group for people with disabilities and their allies. And then, last year we said, "Why does it have to be about accessibility only and why does it only have to be for employees?" So then we opened it up to the community. We had partners and customers and employees all building on the platform, really stretching the platform, bringing us to TDX this year, where builders who really probably never even thought of themselves as social impact developers shipped working Agentforce Solutions for food insecurity, for nonprofit volunteer coordination, you name it. So when we all come together to threshold platform for everybody, not the default user, we're really shipping great technology and boy, did the admin community really show up for this. Mike: Yeah, no kidding. And just as kind of a teaser, next week on the podcast we have the winner of the hackathon to talk through their application. So that was a fun conversation, but let's dive into what some of the things that you did around the hackathon, some of the stuff that you saw and what some of the participants tried to tackle. X: Absolutely, so we had ... so a little fun fact for you is that of all of the submissions for the TDX Hackathon, over 62% of them self-selected to say, "I want to solve for the Agentforce for Good." It wasn't required. And then, 50% of the Agentforce for Good projects were participants could submit in multiple categories, but 50% of those 62% of the projects were only for Agentforce for Good. They solved things from disaster relief to supporting teams post-addiction to food rescue, truly really pushing the boundaries of the Salesforce platform. Mike: Yeah, absolutely. And I always think sometimes that we're trying to solve problems with technology where it's just how do you layer in the technology? But I do feel ... and I was there when they gave the awards, I do really feel that the integral solution for a lot of these works the technology along with the process that is also needed. There's just things you can't do without the technology that when you layer that in, wow, it unlocks a different world. X: Yeah, absolutely. I think that when we are challenged with designing for the edge, you make the center better. And so I'm sure that some of these solutions that put together Tableau and Slack and Agentforce have uncovered some capabilities that maybe we might not have thought about, whether it's for a good cause or not. So I do think it also allows very special kinds of developers and admins to come to the surface as well. Mike: Absolutely. So in your work with the community and around the community, what are some of the events that you go to? What are some of the presentations and content that you put out there that people find valuable and useful? X: Well, I mean, I'm only starting to get into the conference circuit or the community circuit, to be honest with you. I've been kind of- Mike: I like that we call it a circuit. X: Yeah. Mike: Get around the circuit. X: I mean, for me, it's something that I've always been ... and this might be a little too raw, but I am an open book, Mike, so hear me out here. I have never, ever had the ... what's the word I'm looking for? It's kind of like, I guess confidence is the right word. I've never been raising my hand to say, "Oh, I should submit and speak," or, "I should fight to attend this event because this really matters to me." And it's only recently where I've started to get my legs a little bit more to do that. I've published a couple of articles on LinkedIn, building with Cloud Code, using Slack, Slackbot, all these things. And so, I'm just starting to get out there and really getting exposed to the possibilities of our platform. I actually don't know how to answer that question, except for I'd love invites, I'd love to know what the community finds interesting and I'd love to facilitate doing good in the world, doing as much good as we are solving business problems. Mike: Yeah. I think ... I mean, you're also just taking the first step. Some people ... I'll just go, but I'll never present. And I always encourage people, even ... you don't have to go to a big conference or you don't have to present at Dreamforce, but even getting up and sharing things at your local user group, the thing I'm reminded of, somebody needs to see somebody like you presenting. X: Well, thanks for saying that. I'm actually presenting at Connections this week. Yeah. It's my first real speaking session and we're talking about how we've used Slack and Slackbot to scale top level support to our incredible partners. We have over 13,000 partners in the Slack community and we have a lot of different skills that we're using to help our partners go to market faster in a more customized personalized way. And so, we'll be talking about that. But back to the hackathon, the one thing I did want to underscore is that Salesforce has a total of 16 equality groups from Abilityforce, to Neuroforce, to Asiapacforce, to Outforce, BOLDforce. And we all came together for this hackathon and we're going to continue to do that. So likely, you'll see us at Dreamforce and at TDX again. So I would love to have the community just continue to support this and participate and start the platform. I'll say that our community is proving that intersectional volunteer-led initiatives drives higher engagement and faster product adoption than traditional top-down motions. And so, I just want to have a call to participation. If your community or your listeners want to participate, we're definitely open to having more volunteers support and would love to see these at our community events as well. Mike: Yeah. I've already been to a community event MidAtlantic Dreamin', this year in Philadelphia. And I love going to those events because it is all community presenting and really showing their perspective on giving different solutions and tackling different challenges. And it ranges from technical through just giving presentations to your executives and having the confidence to walk into a boardroom and really working through that, not all the presentations are exactly technical. And so, you can learn different skills that way, but I appreciate you bringing that up. That's always very good and good for you getting up on stage. It's okay. I do think sometimes ... I've spoke on a lot of different stages. Sometimes it's the smallest ones that are the hardest because once you get up on a big stage, there's just so many people that your brain can't absorb it. It's crazy. X: Yeah. Well, I guess my little theater session at Connections will have to do, but- Mike: You never know, there could be 200 people there and do a little theater session. X: I know, I know. My goodness. Yeah. What I'm curious about, Mike, from your perspective is why do you think ... Here I am, I'm asking you a question. Mike: Please. X: Why do you think we had such incredible participation? Because it really was humbling the amount of community support we got from folks joining the Agentforce for Good Hackathon. Mike: I thought about that too and I think it's maybe a couple of things. One, I've been a part of either on the team where people around me are planning the hackathon. I've been in the judging room sometimes for a few of the events and going through, watching the submissions and watching the integration of how did they use Agentforce and what was the solution they were trying to solve. I feel like the opportunity for solving things that really provide, I'll call it exponential value to the community, is so great. And it also kind of, in some respects, opens up the creativity because depending on how long you've been a Salesforce admin, how long you've been a Salesforce developer, you might not have all of that kind of exposure into different businesses and different processes. I know for a long time, a lot of hackathon people were tackling things in the medical world and I won't give away ... Well, I can give away the winner because we all know it. But the unique solution that they come up with, you have to think of it's not what is the technology solved for, it's what was the issue and the challenge that provided the biggest gap that technology proved out to do the most good. And I just think after talking with the winner and seeing that solution, I think a lot of people just genuinely found a lot of really satisfying things to build when it comes to solutioning ... I'll call it solutioning for good, as opposed to solutioning for profit. X: Right. Mike: Because then you're like, wow, if we can do this, this and this, because the gaps are always there and they're always huge because nobody ... For lack of a better terms, nobody is maximizing good for profit. X: Yeah. Mike: And so, the thinking of just different things around animal adoption or adoption shelters or food and food insecurity, there's always communication gaps there. And the biggest thing that AI we found and Slack, and some of these other tools, but the biggest thing that AI solves is communication. I was on Amazon the other day just not as a tangent, but kind of ... and they have this new button now you can click and it generates kind of an AI overview of the product and some of the common answers to the questions that they have. And I thought to myself, that gap probably didn't exist, but AI solved it and AI is able to communicate with us. And I think that to me for things that do good and help others, communication is always the key and it's always the biggest gap because communication at scale never happens. X: It's interesting that you say gap and it's making me think of something I'd love to share about our employee hackathon. So we are still doing our employee version of the Agentforce for Good Hackathon and this year, we're adding an AI fluency track. So we have our traditional builder track, and then we have our AI fluency track and you don't have to be building an agent to build something powerful. Mike: Right. X: And so, we want to provide an opportunity for admins and marketers and all sorts of folks to be able to confidently collaborate with AI to give our users agency to drive impact at speed and scale. And so, I think admins do a really great job at understanding human like you were saying the Amazon button and clicking that button as an admin, you kind of know why the user might click it or why they might click the wrong one because of maybe where it's placed or is it accessible or things don't get filled out correctly and you know how to create some workarounds. And so, I think the AI fluency track for us is about really bringing more of that human element back into the AI. So I'd be curious if the community would be down with participating in an AI fluency track where it's like writing ... like it's not about building something. Maybe it's like the most artful skill, Slackbot skill, or maybe it's a killer Gemini prompt or something along those lines. But I think you mentioning the gap really made me think about how I want AI for all technology that empowers everybody, but I also don't want to leave anyone behind. Mike: Right. Yeah. I mean along those lines, I was talking today with my manager and we were discussing something and she's like, "So what do you use?" Because at Salesforce we have a wealth of tools. I think comparative to ... I've talked with other friends and like, "Wow, you get to use that and that and that?" And I'm like, "Yeah, and there's probably more that I don't know about." And they're like, "Geez, we don't even have ..." So that part alone was ... I was like, "Yay. Well, I'm glad I have a wealth of resources." But she asked what I was going to use it for. And I said, "The biggest thing that I try to use AI for in looking at a lot of data on sheets or 10, 15 page documents, what are the blind spots for the stuff that I don't normally pick up?" And that to me was the greatest prompts that I could use. And back to some of your Agentforce Hackathon stuff, I think actually that's some of the biggest opportunities that admins have across their org is we always look at, well, how can the agent improve this or that? Boy, I would just love it if ... how could the agent just look at the record somebody created and be like, "Yeah, Jim, that's not going to cut it." X: Absolutely. Mike: "You left a bunch of fields blank and you know what, those fields are there for a reason. Go back and fill them in." That's what I would want. I would want a little snarcastic clippy coming up and being like, "I think you can do better." X: Yeah. Mike: But it's also not, how do I make myself perform in a greater capacity or increased speed? Yes, that's part of...
/episode/index/show/buttonclickadmin2/id/41523350
info_outline
Salesforce Admins Can Reduce Risk With Proactive Security
05/28/2026
Salesforce Admins Can Reduce Risk With Proactive Security
Salesforce Admins Can Reduce Risk With Proactive Security Today on the Salesforce Admins Podcast, we talk to Kate Lessard, Lead Admin Evangelist at Salesforce. Join us as we chat about why security is the foundation for how data, automation, and AI work together. You should for the full episode, but here are a few takeaways from our conversation with Kate Lessard. Why strong security is proactive, not reactive Security can often feel like the elephant in the room. Everyone knows it’s important, but the amount of work needed to do it right can feel daunting. Thorny problems like accumulated tech debt and organizational inertia can make you feel like you’re better off not talking about it. That’s why I sat down with Kate Lessard. This month, she ran a workshop called Security in Action to highlight the simple steps you can take to strengthen the security posture of your org. She got hands-on with an example org to demonstrate how to use Health Check to identify and fix security problems. It’s all about taking a proactive approach to security, rather than waiting for something to happen to you. Your Health Check score is a security action plan “Health Check is no longer just a list of settings for you to review,” Kate says, “it’s your dynamic risk mitigation dashboard.” Issues are sorted into four categories: high-risk, medium-risk, low-risk, and informational. There is also a status to rate the level of urgency for each issue: critical, warning, or compliant. Health Check is also customizable. If your organization’s security policies are different than the Salesforce Baseline Standard, you can download an XML file and tweak the criteria. You can compare your policies to what’s recommended by Salesforce, and maybe identify some low-hanging fruit. In short, your Health Check score gives you a prioritized list of which security issues need to be addressed first so you can walk into a meeting with stakeholders with a solid plan of action. Helpful reports for user management Security and user management go hand in hand, and Kate recommends some reports that admins can run to help with governance. First of all, the Active System Administrator Report can help you find users who might have more access than they need. This can often be an issue in older orgs where it was easier to give someone admin access than to sit down and solve the permission problem. Kate also likes to run a report on Reports, to keep track of what her users are interacting with the most. As Kate says, it’s all about making a shift towards a security model built around permission set best practices. As admins, we need to find a way to fit security seamlessly into our organization’s business processes, and vice versa. There may be an upfront cost, but it’s all about making security easy in the long run. Listen to the full conversation for more from Kate about security and Health Check. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Blog: Kate Clicks Through It: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, I sit down with Salesforce Evangelist Kate Lessard to unpack why security is really no longer a side conversation for admin, but the foundation to how data, humans, automation and AI all work together. We're going to cover her new security workshop that she led last week, and talk about health check upgrades, permission set security models, and of course how Salesforce admins can move from reacting to security problems to proactively designing trusted systems. We do touch on a little bit of governance, user access, and why secure data is the backbone of every successful AI implementation. So if you've ever wondered how modern Salesforce admins evolve from feature builders into stewards of trust, this episode is for you. And with that, let's get Kate on the podcast. Kate, welcome back to the podcast. Kate Lessard: Hey, thanks for having me back. Excited to be here. Mike: Well, it feels like a lot of relevant things for admins are coming out of your camp lately from True to the Core Deep Dive to last week's security and action. I feel like there should be like a thunderclap after that. Security and action. Dun, dun, dun. Kate Lessard: Yeah, absolutely. Data and access. Mike: Right. Mm-hmm. Yep. Like a comic book hero. And then you got to do the comic book land. Anyway, totally off topic. Look at that. 30 seconds into the podcast, we're already in comic books. People- Kate Lessard: I mean, I am here for it. Admins are superheroes, especially when they are securing their orgs. Mike: We are. So let's talk about what that first workshop was, and what you covered, and how it went. So tell me a little bit more, tell everybody a little bit more about what that workshop was. Kate Lessard: So in the Security and Action workshop, we were really focused on data and access and using HealthCheck as our guide. So essentially we started off with some security basics knowing that much of our audience was at different levels. So we had people joining us that are brand new admins that are working on their first certification. And then we had people joining us that have been in the ecosystem and in a Salesforce role for more than 10 years. So wanted to make sure that we had a strong foundation and that we had reviewed some security foundational knowledge base points for everybody to get on the same level. And then we got hands-on in an org that just didn't have a great health check score and had some security concerns that we went through, and hands-on adjusted together until we improved our score, we better secured our org, and just increased our security posture. Mike: Cool. I didn't hear Agentforce in there and we're in the age of AI. Kate Lessard: Yeah. So I do think we didn't specifically touch on Agentforce in this first workshop. We really wanted to focus on that strong security base where it applies to all of your org. And that might mean that it applies to your users, both human and agent, but really focusing in on that prime security. And then we do have plans in the future to enhance this and take this to the next level and really expand on some of those agentic guardrails. But I think that it's just really important to have that foundation first. Mike: Yeah. I mean, all of it bases around good, clean data and secure data. Kate Lessard: Absolutely. And you're not going to have a successful Agentforce implementation without that. That is the foundation point that's going to set your organization up for success using AI. Mike: I did the Agentforce Now workshop that week, and I always emphasize the importance of filling in description and metadata fields as well. Kate Lessard: Yes, hugely important. Mike: So one of the things, it's really cool the workshop pivots all around HealthCheck. And I know having worked some of the Q&A, people were like, "Wait a minute, I thought HealthCheck went away." Kate Lessard: Yeah, HealthCheck has not gone away. It is free for admins to use and it actually recently got some pretty impressive and exciting upgrades that we spend some time on in the workbook. So it's no longer just a list of settings for you to review. It is really dynamic and serves as your risk mitigation dashboard. So you can do things like not only configure the specific settings for your org security, but you can also set up email notifications for system admins or anyone who that you would like to receive notifications when your security score changes. So maybe members of your IT team, your security team, if you work with a governance team, making sure that they get notifications because as you know, security is a team sport. I think the coolest thing that we do in the workshop and that I really want everybody to be able to take advantage of is you can actually export the standard Salesforce baseline and customize it to your own organization's security criteria. So if you have different security criteria like maybe your password policy has a minimum of 11 characters at your organization, the Salesforce baseline has eight characters. That is a change. And you want to compare your Salesforce security settings against your own organization standards, and you now have the ability to do that, which is just incredible. Mike: Yeah. And it's not like the way that you walk through it, it's not that daunting. I mean, is it an XML file that you download? Kate Lessard: Yeah, absolutely. So you don't have to start from scratch, which is the really nice thing. You can actually export the standard Salesforce baseline. It is an XML file. In the workshop, we walk through what the download looks like and how to actually make the edits. It's pretty easy to read. And even in the file, there's a link to help notes so that you know exactly what types of formatting to use when you're editing the XML. And then you can just save it as your custom, re-import it, and set it as your default. Mike: Yeah, I thought that was really cool. One other thing that you dive into is, and of course it's set up because it's a workshop, but number of admins that... Kate Lessard: Oh yeah. Mike: Or number of people that have the Salesforce admin profile. Kate Lessard: Absolutely. So everybody's favorite admin horror story that you log into an org for the first time and take a look and realize that you have just a completely disproportionate number of system admins and folks that are over-privileged just because it was easier to give them that permission than to really sort through a proper permission set model. And in our org, we have the very scary informational security setting letting us know that we have 100% of our users as system admins opposed to the typical 5% that you see in the standard baseline. So we address that and we talk about who really needs admin permissions. Mike: So I mean, let's expand upon that because we've had that idea of delegated admins as well. So they wouldn't have a system profile or a system administrator profile. Kate Lessard: That is correct. Yeah. You might have your delegated administrators that are taking certain tasks for certain teams, but that doesn't mean that they have the full system admin profile. It means they're a delegated admin and they have certain admin responsibilities that they take on, but it does not give them the keys to the castle. Mike: Very important. Very important. With HealthCheck, I think it's always one of those big discussion areas because there's different levels that we call out. We have critical. What are some of those stages? Non-critical, informative. Would you as an admin getting started approach HealthCheck as maybe it's like a topic we should bring up in governance? Kate Lessard: I absolutely think it's a topic we should bring up in governance. And it really is set up well to help you prioritize. So if you're going into a governance conversation or maybe a meeting with your security stakeholders, and you want to present different actions that you would like to take or approaches and settings that you'd like to address in your org, it's broken down by high risk, medium risk, low risk, and then informational. And then in each of those different categories, we have it called out what is a critical status, what is a warning status, and then where you're compliant. And so you might want to prioritize your high risk critical items, and then you could share that information of what your value is versus that baseline so that you can provide a really strong recommendation about where to take action to better secure your org. Mike: That makes sense. You've been doing a lot of security stuff lately. I mean, you've had some workshops. Outside of this, we had the True to the Core Deep Dive. What are one of the areas that almost consistently comes up as a theme that really maybe this year admins could address the most? Kate Lessard: That is a really great question. I think we've seen it in the True to the Core Deep Dives that we've had. We had one recently on security and user access. And I think those are two things that go hand in hand and something that comes up consistently. It was also a major discussion point at our admin day of security before TDX. We had a day zero event with several admins and that was just a common conversation as well. Really making that shift towards the best practice of having a permission set led security model, what that looks like for your organization, how to get there, what tools are available to help. I think that things like user access summaries, which we dove into in the workshop as well last week. There are just so many tools that can help admins better move to that permission set led security model. And I mean, I don't have to tell you that security and user management really go hand in hand, and oftentimes are interchangeable topics about how we can better secure our orgs and create a better user experience as well. Mike: Absolutely. And being proactive on topics that come up, one of the things that you cover that I really enjoyed, this is also, this speaks to how long I've been in the ecosystem. It wasn't until I joined Salesforce that I think you could do really in depth user reports, basic reporting was available, but one of the things that you create is the active system administrator report. I'd love to know, I mean, you were an admin out in the world more recently than myself. Were there reports that Kate always created to help her understand her users outside of just if they had a license or if they didn't? Kate Lessard: Yeah, absolutely. I think that the active system administrator report is a great one to call out, but it is looking specifically at users that have that system admin profile. And you might have other concerns about your users or things that you're checking up on them and what they're doing and what they're using. I loved to create the report on reports to get little snippets and information into what reports my users were using. I found that to be really helpful. And also there is a great Salesforce Labs report package about different administrator reports to help you get a little bit more insight into your user activity. So I think that something like that is really helpful if you don't have advanced security options to pinpoint that user activity for you. Mike: Yeah, that definitely makes sense. In addition to knowing when people log in and when they don't, it's also what are they doing? Because I remember I did an implementation and they were adamant about having a login component on a dashboard. And I had one user that I, to this day, I think all they did was wake up, check their email, and then log in and out of Salesforce because they were always head and shoulders above everybody else. They'd have 300 logins for the week. Kate Lessard: Oh, interesting. Mike: I'd be like, "Are you just logging in, updating a field, and logging out?" And they were very dodgy about their answers, but I think that's what they were. I think that's what they were doing because they liked being at the top of the board. Kate Lessard: That is hilarious. Mike: And I always had to reframe the conversation with executives of, we have to look at the entire dashboard, not just logging in because it could be meaningless. Kate Lessard: Right. They're just like logging in, logging a single activity, logging out, coming back in. Mike: Refreshing a dashboard, refreshing a report, and then that's it. Kate Lessard: Oh my goodness. Well, and then the other concern is where are they logging in from, if we're talking about dodgy behavior. When I was an admin out in the wild, one of the stories I like to tell about why security is so important. I worked for a mental health hospital system, and as you can imagine, we dealt with a lot of PHI. And I noticed some funky login IP ranges, and it turns out that someone had given a local college student access to their Salesforce credentials and was having them log their activities for them. Mike: Uh. Kate Lessard: Yeah, I know. And thankfully they didn't actually have access to any PHI because we had a good security model set up. However, obviously huge concern and something that Salesforce was able to help us identify because we were able to see those different login IP ranges, spot what was going on, and address it, and prevent it from becoming a big security concern, which I feel like that's really one of the big takeaways that I... If you asked me what I wanted people to take away from this workshop, it is to shift from being reactive about your security and just waiting to become a statistic to really being proactive and getting ahead of things and making sure that your org is secure and that you are addressing those risks versus waiting for them to happen to you. Mike: Yeah. I mean, to that point, I often look at things that are security that don't sound like security. And one of them is process management. So to me, that sounds like there's a complete separation between what executives expect in Salesforce and what the actual process is. Because the process, the technology should support the process that's needed, and here it sounds like it's burdensome to the process. And that to me is a security risk. I mean, if you don't have the process ironed out so that it feels natural, every action has a deliberate and equal technology component, then the security's going to fall apart. Much like a Salesforce admin shouldn't learn about a person being separated from a company four days after they've left the company. They need to be part of the off-boarding or onboarding process. They don't need to know why, but they need to make sure because that person should or should not have access to that system. Kate Lessard: Absolutely. I think that it is certainly a role of power and responsibility because you are getting that notification, best case scenario, to freeze that user. And again, you don't need the why or the explanation, but to have an actual process in place to be able to continue to secure your org so that after someone's left an organization that you're able to say, "Yes, this is still secure, this is not a risk." And you know you're not maybe going through a full deactivation process right away, but that ability to freeze that user and be able to take action is just so important. Mike: Absolutely. Well, Kate, I appreciate you coming on the pod and helping admins keep all of their data secure and their org secure. Kate Lessard: It is my pleasure. I mean, I love to talk about security any chance I get, and I think that we have so many brilliant people on our security team at Salesforce who I'm always learning from and happy to partner with. And this workshop was just a real labor of love, and I'm so happy that so many people were able to attend last week and glean the information and be able to take that and go use HealthCheck and better secure their actual orgs, not just our fake, very poorly-secured org that we fix and adjust in the workshop. Mike: Yeah. I mean, it works because you can spot things very easily because I think in the real world, sometimes things aren't as obvious as they should be. I know the biggest question that we can answer, and we're doing it at the end on purpose, will there be more of this content? Kate Lessard: Yes. So I can't give you a date, but we are planning to continue this on a monthly basis. I do think that there are plans to expand the content as well. And you'll also maybe catch us doing these workshops hands-on at some community conferences over the next year. Mike: Ooh. See, even more of a reason to go to community conferences besides the fact that they're super fun and held at really interesting locations. Kate...
/episode/index/show/buttonclickadmin2/id/41444150
info_outline
Can AI Help Salesforce Admins Build Apps More Efficiently?
05/21/2026
Can AI Help Salesforce Admins Build Apps More Efficiently?
Today on the Salesforce Admins Podcast, we talk to Nick McOwen, Senior Salesforce Administrator at Alpine Intel. Join us as we chat about his path to Salesforce and the TDX workshop he gave about the development lifecycle, sandboxes, and data masking. You should for the full episode, but here are a few takeaways from our conversation with Nick McOwen. Nick’s unconventional path into Salesforce Nick is the second touring musician I’ve had on the show this month—I promise I’m not raiding a recording studio somewhere just to find podcast guests. He was waiting tables in between gigs when a guest called him over to ask if he might want to try something different. A chance encounter turned into an entry-level job and, eventually, an admin certification. I caught up with Nick fresh off his workshop at TDX, where he shared how he uses sandboxes and data masking to build new agents and apps for his organization. Learning better development habits and sandbox management They say every warning sign has a story behind it, and the same goes for best practices in Salesforce. Luckily, Nick had Salesforce MVP Kelly Bentubo around to show him the ropes. He learned about the importance of consistent naming practices, managing user expectations, and having a structured deployment process instead of building in production. Recently, Nick was tasked with building a new recruiting app for his organization. He was able to spin up a sandbox with an app they had already built and, using that foundation, quickly reconfigure it to meet the new requirements. Once the app was in good shape, Nick moved it up to a staging sandbox environment for testing. There, he could copy data over from production and use data masking to keep everything secure. Once everything was thoroughly vetted, it was finally ready to be deployed into production. How Agentforce Vibes can help admins collaborate with developers Recently, Nick’s been taking advantage of Agentforce Vibes to work more closely with his dev team. He was able to write an Apex class and, while the code isn’t perfect, he was able to go through it with a developer to learn what was working, what wasn’t, and why. “It was a great launching point,” he says, “something that would have taken days was written in an hour.” For Nick, the most important thing admins need to do to get the most out of Agentforce Vibes is to learn the basic underlying principles of Apex and coding. It’s just like using a calculator—you still need to have some way of knowing if the answer you’re getting is in the right ballpark. Make sure to listen to my full conversation with Nick for more on sandboxes, data masking, testing, and why AI is a new opportunity to grow. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Salesforce 360 Blog: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, I'm joined by Nick McOwen, which means now I've started a trend of having former musicians who become Salesforce admins as guests. Anyway, Nick is going to show us how he went from touring in a band and waiting tables to leading Salesforce projects that involve sandboxes, data masking, deployment processes, and AI assisted development. We talk about the early build it in production and hope for the best admin days, how Agentforce Vibes is changing the way admins approach coding and why today's Salesforce admins are becoming orchestrators of systems instead of just builders of features. There's also a little Apex talk and surprisingly deep analogy involving semi-trucks and AI because you know my analogies and a confirmation that developers will always think their code is prettier. So hit subscribe, share this episode with your favorite Salesforce admin and let's get Nick on the podcast. So Nick, welcome to the podcast. Nick McOwen: Hey, Mike, it's great to be here. Mike: Well, I was introduced to you through a couple product managers at Salesforce that are always on the hunt for finding good Salesforce admins and you were also presenting with them in the admin track. So before we get to that, let's learn a little bit about Nick. How did Nick get started doing Salesforce and be in the ecosystem? Nick McOwen: Well, it was a complete fluke. I had been touring in a band and we just kind of stopped playing music and I was waiting tables at a country club and one of the members was, "Do you want to keep waiting tables?" I said, "Not particularly." So got me a job doing the entry level position at the company and I realized I didn't want to be doing that either. So I went to the developer at the time who was also the CIO and the CTO and said, "I'm going to learn Salesforce." And he kind of looked at me in the way that an adult might look at a middle schooler who said, "Watch me hit this three point shot." And I eventually learned it after a few misses, but they decided to take a risk on me and many, many years later here I am working with Agentforce agents writing a little bit of Apex here and there. And yeah, it just didn't intend to get here, but here I am. Mike: I mean, people are going to think I went down and raided the local music festival because not but a couple weeks ago I just had Adam Stark on who was also a musician that became a Salesforce admin. So apparently it's a trend. Nick McOwen: Yeah, that's the track. That's the track line for most musicians, I guess. Mike: Yeah. I mean, let's see. Should I be like an '80s metal hair band or an admin? Nick McOwen: Oh, that's a tough call. Mike: I know, right? Let's talk about some of the stuff that you presented at TDX because you covered a lot of things in terms of development life cycle and sandboxes and I felt like you were in that area for stuff that admins are interested in and kind of that next level of their career building apps. Nick McOwen: Yes. It was a talk that focused around seeding sandboxes masking data and my experience with that and it's funny because prior to the presentation at TDX, shout out to Allie and Sam, I was talking to them and they were running a booth and they were saying how many people that they'd talked to who didn't use sandboxes or didn't quite see the value in sandboxes. And so I realized that this talk probably had a lot more impact than I thought it was going to be. My particular use case was how we had rebuilt an app that our recruiting team was using for our field agents and how we spun up a sandbox specifically for it, rebuilt everything using the foundation of the app that was already there and got to test in a secure environment, move it up to our staging environment and then production. So it was an interesting experience and an eye-opening one too, getting to talk to everybody. Mike: I mean, to be fair, I don't think when I started as an admin, I really had to pay attention to what my organization's life cycle for app development was because you're kind of, and in this case, and I don't know about your company, but you're kind of a small scrappy unit and maybe they got Salesforce for like 20 or 30 people. I mean, nine times out of 10, I was just building stuff in a dev org to make sure it worked and then literally just not even deploying it, rebuilding it in production and Monday afternoon being like, "Okay everybody, so there's a new field on the account." I think back to those days of like, "Wow, what was I doing?" Nick McOwen: 100% identify with that because that is exactly how I more or less learned when we were first starting out because like I said, it was just the developer and then me and we had to get things done because it was a startup environment, things move fast and so it was tested in the sandbox and then move it. And I didn't want to bother him so I wasn't using change sets and so I would make a field or whatever he asked me to make and then rebuild it in production when nobody was working and just, yeah, it was loose and fast. Mike: I mean, nine times out of 10 when I was building things in my dev org, I wasn't paying attention to naming. I was just trying to like proof of concept and so I never wanted to package anything up because I didn't want to move it over because I was like, "Well, when I get to production, that's when I'm going to make the names nice and shiny." Nick McOwen: Yeah, I remember we brought on Kelly Bentubo, MVP. She sat me down and was like, "All right, we're going to start fixing some of these habits," which totally warranted. And I mean, she was great at helping me understand the value in the consistent naming, the process aspect of everything. So she really elevated my abilities and my use cases and everything. Mike: So as somebody that just started, obviously went from waiting tables, which by the way, 20% tip is not excessive. It just means that you got everything good and on time, you should tip more. Thank you for attending my waiter tipping TED Talk. Just getting into this and thinking, okay, I guess these are bad habits. I bet you had the same bad habits I did. What was the first things that she kind of tackled? You mentioned some of it was naming. Did she help you set up sandboxes and kind of a deployment life cycle? Nick McOwen: So we had the sandboxes and she came on at the time when the dev team was starting to grow and the company itself was starting to grow. So we needed more organization and she kind of sat me down and helped me with these processes, like how to move it up through the environments, the value of making sure that everything is tested within the sandbox environment and then user management, which I think is something that I relied on when I first started. It wasn't necessarily that I knew what I was doing a lot, but I knew how to put on like a waiter attitude so to speak. And if somebody asked for something and I didn't know what I was doing, it was just that kind of like, "Oh, I didn't burn your steak, but let me make this better for you." And so through talking with Kelly, I kind of like refined how to shape expectations for users. If I couldn't figure something out, then she would kind of help coach me through the deployment issues or level setting and expectations with users, which that was a big one. Mike: I mean, I think that also we've talked about order taker admins before and it's very easy to sit across the desk from somebody and listen to them be like, "Oh yeah, this is totally easy to do," or, "I have no idea what I'm going to do." And then say, "Yep, let me just work on it and kind of forget that maybe there's bigger things in the organization they need to pay attention to as opposed to everything that comes in front of me right away." Let's get into that recruiting app that you built. Was that your first kind of get your hands dirty in the org per se, building something new for the organization? Nick McOwen: That was my first project that I led from start to finish. I got the specs from the higher ups. They said, "This is roughly what they're asking for, figure it out." And immediately went to the recruiting manager and said, "Okay, let's figure out how we want this to look." And we had an app that was already built and I knew that the foundation of that app was great. And so it was really a matter of building off of those blocks and reconfiguring it, re-imagining how it could be used and then future-proofing it so that it could scale as the business grew. And it was definitely a lesson in organization and communication, which is good. Always want to keep learning and growing. Mike: Yeah. So let's dig deeper into that organization communication. What organizationally, do you mean the way that you organized stuff, understanding your company's organization, a little bit of both? Nick McOwen: A little bit of both. For me, it was keeping track of everything that was getting worked on, making sure that we were trying to document the change in processes and communicating those changes to the business side so that they would know, you might not interact with this functionality, but this is what's going to happen to get you this result. And that it sounds pretty straightforward, but I think that can get lost sometimes on people, especially in larger projects where it seems peripheral and intuitive, but it might be what's intuitive to someone who built it might not be for the person that's going to end up using it. Mike: Yeah, which leads into communication. I always feel you can't communicate enough. I feel with every project I've ever worked on, you always learn more about communication. What did this first jump in kind of give you in terms of maybe some best practices that you follow now or stuff that you learned really quick? Nick McOwen: I think the accountability aspect maybe was if you presented an idea and it's kind of your child and you say, "What do you think about this, huh?" And then the end user going, "No, that's not going to work for us." There's like a little bit of an ego bruise where you think, "That was a really good idea though." Maybe it wasn't, but then kind of absorbing that critique and then using it to build something that you know is going to work and not being stuck ruminating on the fact that your idea didn't get accepted, that was kind of a lesson because prior to that it was somebody else's idea or I was building something that was given to me. It wasn't as personal, but that was the first time I got told, "Hey, I don't like that." Mike: Yeah, that can always... It's funny because you almost want your cake and eat it too, right? I don't want to take the rejections personal, but you 100% take the wins personal and it's like you can't have both. You really do though because it does feel like, "Well, no, I know how this thing works and I promise this will really work for you, but you're just not seeing it." And that can be the frustration that I always ran into. Nick McOwen: Exactly. And then even after I built it, my bosses came up and they said, "What is this object?" Oh, well, I made that because that was giving the solution to the end user and they said, "Hmm, well, why didn't you do this?" I was like, "Ah, I don't know why I trusted my gut and stuck to it." Mike: Right. Well, this is the way it is now. Let's talk about, I feel like that's always the first part. Your path into Salesforce, they always talk about individuals that grew up and like I was in high school in the 90s just to date myself, but I remember in middle school there was no internet and when I got to high school, you could go to the library and check out the internet for 15 minutes on a computer. I bring that up to say, obviously now everything's online and I don't know how I even got anything done in 15 minutes with dial-up modems back then, but apparently I did. You kind of came in to the Salesforce ecosystem at about the same kind of inflection point with AI in that there's really best practices and development aspects to learn, but then, oh, hey, by the way, now we've got a lot of new stuff coming like Agentforce Vibes. How did you kind of balance those two that learning path? Because I think it would be very easy to, "I don't need to learn this. Vibes can do it for me." Nick McOwen: I think that's something that is incredibly important in this time with the ubiquity of AI and how many people are using it. I think people need to become literate in AI. I don't mean learning how to program large language models or even write agents, but just understand how it functions. And in the context of Salesforce, I remember I first started getting into the predictive analytics for Einstein and thinking, "Oh, this is super cool." We didn't end up going that route, kind of kept my eye on everything going and then Agentforce came out and started learning that. And it feels like every couple weeks something's getting rolled out, which is exciting because there's always something to learn and I don't know, I would hate to be stagnant and have to do the same thing over and over for 20 or 30 years. So it's a blessing and a curse to have a product that constantly evolves for you to learn. So it's exciting. Mike: I think you covered Agentforce Vibes in your TDX workshop. What have you done in Agentforce Vibes? Nick McOwen: Shockingly, I made an Apex class and a test class that- Mike: Hey. Nick McOwen: If you had asked me... Mike: It's kind of what it's there for, right? Nick McOwen: Yeah. The musician that could barely string sentences together is all of a sudden writing Apex classes and test classes. I mean, if you asked me even a year ago, "Hey, Nick, would you think you could write an Apex class?" I was like, "Yeah, give me a couple years so I can write it line by line over days." Mike: Sure. Let me put an ad out on Craigslist and see if I can get a developer to do it. Nick McOwen: Yeah. I would've laughed if somebody said that and then all of a sudden this product comes out, which actually Allie and Sam, when we were doing our ... We did a webinar for the sandbox seating and data masking and they were talking to me about the Agentforce Vibes and I said I hadn't really looked into it. And so after that talk, I started to and that kind of encouraged me to look into the foundations of Apex and coding languages. So it wasn't just learning Agentforce Vibes and how to use it. It was also getting a base understanding of the underlying principles. And so through that journey, I was able to use Vibes to write an Apex class and then I could take that class to the developers and then the developers would look at it and go, "Well, my Apex is better, but we'll use yours." And corrected a couple things because... Mike: Wow, you really nailed the developer intent like mine's better, but we'll use your... Nick McOwen: It could be flawless code and still, mine's prettier. No, I do. I enjoy working with all those guys and they were great helping walk me through everything and showing me, "Okay, this is where Vibes made this. This is the intent but it doesn't quite hit. Here's how we're going to fix that. Walk me through." And it was a great launching point. Something that would have taken days was written in an hour. Mike: Oh, wow. Nick McOwen: We're not talking thousands of lines here. It was- Mike: No, but still. Nick McOwen: 150. Yeah. And then the test class on top of that, it saved us a ton of time. Mike: If you threw me in jail and said, "You're out when you write 150 lines of Apex class," I'll be like, "What's for dinner?" Nick McOwen: I've been be here for a while. Let me get some furniture. Mike: We're going to be here for a few years. Do I get a shave every now and then every few weeks to keep me from looking scruffy? I mean, I was so impressed getting the admin track ready for TDX this year with the submissions that admins had. I watched them demo vibes and just use the planning tab and really kind of have vibes spell out like, "Here's everything you need to do and here's kind of an action plan." And I felt like natively some people had that, but for a lot of people, wow, this was really helpful because I had this monster of an application but I didn't know where to start and it wasn't just, "Oh, I have to vibe code this super hard application." It's like, "No, this is actually just helping me get my steps down." Nick McOwen: I'm excited for the future of the vibes and Salesforce integration, specifically with your organization's metadata and being able to find, where are these fields referenced? What would happen if I did this? I think when that becomes a reality, it could be something that's really cool. And I imagine that's something that's possible if it's not already. I'm not a vibes connoisseur just yet, but I think something like that would be incredibly beneficial. How many times have you tried to figure out where an error is coming from...
/episode/index/show/buttonclickadmin2/id/41364320
info_outline
How Headless 360 Helps Admins Bring Salesforce Anywhere
05/14/2026
How Headless 360 Helps Admins Bring Salesforce Anywhere
Today on the Salesforce Admins Podcast, we talk to Khushwant Singh, SVP of Product Management for the AI Application Development Platform at Salesforce. Join us as we chat about what Headless 360 means for admins and how Salesforce is building towards a future where setting up complex systems becomes a conversation, not a checklist. You should for the full episode, but here are a few takeaways from our conversation with Khushwant Singh. What is Salesforce Headless 360? If you’re not a developer, the idea of a “headless” platform may sound a little spooky. My guest this week, Khushwant “Khush” Singh, is here to tell us what it all means and why Salesforce Headless 360 represents a step forward into the future of the Agentic Enterprise. As Khush explains, in a headless platform the frontend is separated from the backend. If you have a pixel-perfect website that you love and you want to use Salesforce applications, or visualizations, you can do that without having to change anything. If your organization relies on Slack, your users can interact with Salesforce data and agents in the chat instead of having to open another browser tab. “Headless 360 is about bringing Salesforce to wherever you need it,” Khush says. Headless 360 builds the foundation for the Agentic Enterprise One of the main reasons for Headless 360 is to rethink the Salesforce stack for AI. “For the last 27 years, those layers have been put in place for a human,” Khush points out, “so how do we open this up for both a human and an agent?” To do that, Khush and his team are translating the entire platform into a set of model context protocols (MCPs), skills, and metadata that can interface directly with AI agents. This makes conversational tools within the platform, like Setup with Agentforce, even more powerful, while also allowing you to use any other outside AI model you like. And while these changes have the biggest impact on vibe coding for now, Headless 360 lays the groundwork for a future where we build agents, create applications, and configure Salesforce through a conversational interface. No coding required. Why trust is even more important with the rapid pace of AI AI is quickly transforming the way we work, and the conversation can become centered on how to move fast. But that makes trust, quality, and governance even more important, Khush explains. Headless 360 is built with trust as a core feature, so you can move quickly without making compromises. There’s much more from Khush about Salesforce Headless 360 and what it means for admins, so be sure to listen to the full episode. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce 360 Blog Post: Salesforce Admins Podcast Episode: Salesforce Help Article: Salesforce Admins Blog Post: Trailhead: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: This week on the Salesforce Admins podcast, I'm joined by Khushwant Singh to talk about what happens when Salesforce stops just being a destination and starts becoming part of every system your business already relies on. That's right, from Headless 360 to Setup with Agentforce, Khush breaks down how admins are moving beyond configuring features and into orchestrating how data, AI agents, automation and people work together across the business. Now, we're going to dig into what Headless actually means for admins, why governance and trust matter more than ever in an AI first world, and how Salesforce is building toward a future where setting up complex systems really becomes more of a conversation than a checklist. This episode is really about systems thinking, how admins can connect experiences without forcing teams to rebuild everything from scratch. So if you've been trying to figure out where AI fits into your role or how your role is evolving alongside it, I promise you this episode gives you a practical look at what's already happening and what's coming next. So be sure to hit that subscribe button. Share this episode with your favorite architect or admin friend, and let's get Khush on the podcast. So Khush, welcome back to the podcast. Khushwant Singh: Thank you for having me back, Mike. Mike Gerholdt: I say that because I had to look and it was 2022 back in the day you were talking Experience Cloud when we last had you on. You've since moved on, moved up, had your hands in a lot of things, did an admin keynote with me. So let's start there and let's kind of catch everybody up on some of the products you've been working on, and then we're going to talk about the newest one that we launched at TDX this year. Khushwant Singh: Yeah, happy to. And again, great to be back. I think back in the day, started obviously with Experience Cloud. Experience Cloud really was building a product that is external-facing and again, it's built on the same platform. And again, as we looked at the product portfolio, it just made sense to start to bring things together. So we've actually sort of rolled and morphed into the application development platform, obviously within the platform organization, everything pro-code, low-code, desktop, mobile. And so there've been quite a few products leading up to here, anything from our delivery on mobile, mobile offline, the Salesforce mobile app, all the way to Agentforce Vibes, the Agentforce Experience layer, obviously a fair amount of work on the Headless side of things. So yeah, it's been quite a journey to this particular point, Mike. Mike Gerholdt: Yeah. I mean, it's actually hard to point out things that aren't admin-friendly that you've had your hands on because you've always kind of championed everybody being able to build on our platform. So let's talk about TDX this year because we launched Headless 360. Khushwant Singh: Oh yeah, we did. Mike Gerholdt: By the way, in case you forgot. Khushwant Singh: Yeah, it was a massive announcement. I think, look, as you think about what's happening in the market right now, you've got customers engaging through many different surface areas. We've got customers on ChatGPT, on Claude, on Gemini, Teams. At the same time, you've got developers also using a variety of different tools, different IDEs, different coding agents. And so I think Headless 360 really is a testament to acknowledging this change in the market in the industry and trying to meet the customer where they are. The customer being an end user or the customer being a developer and admin, just meeting them where they are in whichever tool, whichever interface they might be. Mike Gerholdt: Yeah. I mean, do you recall ever a time in technology when it was, I hate to use the euphemism, but like the Wild West where I feel like there's so many choices for what you can do just in regards to AI? Khushwant Singh: There is. And there is obviously a good side to it. There's also a downside to it. The good side is really your [inaudible 00:04:31] for choice. There is innovation that is helping you drive productivity day in, day out. The downside of this is you just have to be really careful about quality, about trust, about governance. And you have to ensure that while you're going fast, you're not sort of compromising on that trust aspect of things. And that's what we're trying to do here at Salesforce and just trying to be a little bit more thoughtful in our approach over here. Mike Gerholdt: Yeah. So my developer friend, Josh Burke, who comes on the podcast, usually has to explain things to me every now and then. I think developers are very familiar with the term headless. Khushwant Singh: Yes. Mike Gerholdt: I'm not sure outside of Headless flows, which I know what those mean, I'm not sure admins could understand or could confidently buzz in on a game show and answer what is Headless 360. So can you help them kind of understand what is Headless 360 if they had to tell their manager? Khushwant Singh: Yeah, sure. So one way to think about it is, let's say you're a customer and you've got your own website, it's built out of React, it's not built on the Salesforce platform, but there is so much of your goodness and investment that you have within the Salesforce platform that you want to be able to easily reuse in the context of this other website that you have, portal that you have, mobile application that you have. All of that may not be built on the Salesforce platform. You want to be able to reuse your investment seamlessly. And so Headless 360 really is about how do we empower you, the customer, to be able to easily use those assets, those investments within the Salesforce platform in that surface area that you've got that may not be on the Salesforce platform. So it's really about giving them that flexibility so that they can easily reuse their investments in all of these various surface areas and just meet their customers wherever they are versus bring their customers into Salesforce. Mike Gerholdt: Yeah, I think it makes sense. I mean, I was a customer way back in '06 when I remember it was like transfer and move to Salesforce, and everything was "have it on the platform, have it on the platform." And then shortly thereafter it was kind of like almost the realization that we had was, "Well, it's a lot of friction to constantly move platforms. What if companies just integrate and we become that connector to everything?" And I feel the same way with this. Khushwant Singh: 100%. And you know what? Look, I think the one example is if you go into the Williams Sonoma website, right? Williams Sonoma website, it's not built in Salesforce, but if you use their sous chef AI, I think... I'm not sure what they call it, but- Mike Gerholdt: I'm sure it's something really, really snappy and small. Khushwant Singh: Olive. Olive. There you go. Mike Gerholdt: Olive. Yeah. See? Khushwant Singh: A snappy name right there. That agent is an Agentforce agent, right? So that's on example where Salesforce is being used in whichever head that you'd like us to be in. And so in this particular case, the Salesforce agent is in the Williams Sonoma website. So that's just one example where we are trying to make it easier for this integration to happen through the Headless 360 platform. Now that's one view. And that's the view, I would say, an easy way to try to understand what it means for Headless 360. Mike Gerholdt: Yeah, because then as an admin, as a developer, you're not having to sit down and say, "Okay, well, if we move to the Salesforce platform, we have to deprecate everything that we just spent the last two years building on our site." Marketing's like, "No, no, we just got the site pixel perfect the way we want it. Now we can integrate best of everything that we want." For admins who aren't thinking of maybe websites, how does Headless 360 play into other integrations that they may talk to IT or internal stakeholders about? Khushwant Singh: Yeah. So for example, when admins are chatting with their other internal team, they may have a custom application that they already have in place and they want to be able to have Salesforce data, Salesforce visualizations surface in whichever application they may have already, right? That's one example. So again, it is about bringing Salesforce to wherever they may already have. So for example, an admin may have a business application that already runs within ChatGPT. So ChatGPT, for example, it allows companies to create bespoke applications. And if you want to be able to integrate Salesforce into that application, that's doable as well. It doesn't have to be in ChatGPT. It could be in any other existing non-agentic application as well. But it's about that, look, you don't have to rebuild everything. Wherever you are, you've got an existing application, we'll find a way through Headless 360 to be able to bring Salesforce to it. Mike Gerholdt: Yeah. No, it totally makes sense. I mean, in that way we can really think about how are we making sure that the departments have the software and the tech that they want, but also access to all of the information that they need to do their jobs the best. Khushwant Singh: Mm-hmm. Mike Gerholdt: So Headless 360 is a big thing we talked about. I also know I saw you in some True to the Core and True to the Core Deep Dive stuff. Was there some features you were talking about in that that possibly play into what admins are going to be working on for the rest of this year? Khushwant Singh: Well, yeah. I mean, a few aspects as we think about Headless 360 that would be of interest to admins. So first and foremost, you've all seen the stack diagram of the Salesforce platform. You've got those four layers. You've got your layers of engagement, you have layers of context, you've got the sort of our data foundation, et cetera. So you've got those layers that make up the Salesforce platform. For the last 27 years, those layers have been put in place for a human, right? And now what the product teams and the engineering teams are doing are going back and saying, "How do we now open this up both for a human and an agent so the agent can do that work on behalf of the admin when the admin sort of wants a task to be done for that particular reason?" And so what that means is that we're translating our entire platform into a set of MCPs, a set of skills, a set of metadata that is grounded so that these agents can do the work on behalf of the admins. So one good example is the setup with Agentforce, which is so far so good. We've gotten great feedback on it. Now behind the scenes is a set of skills, a set of MCPs that are firing off to be able to get their tasks done. And so that's one thing that I think our admins can look to benefit from. The other piece is I know we have some admins who are familiar with some of these IDEs that we have out there. They use Cloud Code quite a bit. They use Vibes quite a bit. It may not be the entire admin audience, but for those audiences, again, the ability to be able to use all of these MCP skills in these development environments is another thing that our admins can really look forward towards. Now the other thing that we've also worked on is, over the last few months, the foundation for these skills, these MCPs, this Headless 360 foundation has been very sort of focused on how do we unblock these pro-code use cases, whether that's in Cloud Code, whether that's in Vibes, et cetera. But this is the same foundation upon which as we speak in safe harbor, I share that we are building what we call our no code offering as well. So how do you go about in an unintimidating sort of type of interface, have this conversation with an agent, which behind the scenes is interacting with that same Headless 360 layer to help you build out your agents, your applications, to help you augment your existing agents and applications, help you set up provision, configure Salesforce from a zero to one point of view, because we all know how setting it all up is quite a taxing task for our admins? So that's the potential I see for Headless 360 for our admin personas. Mike Gerholdt: Yeah. I mean, to go back on some of the stuff that you were mentioning, I remember the first time I thought, "Hey, I could have ChatGPT rewrite ChatGPT's instructions." And I remember thinking like, "I'm so far ahead of the curve, man. I'm making AI rebuild AI." And then it was only a few weeks later that Cheryl was demoing Setup with Agentforce and I thought, "Oh, that makes so much more sense because all of the things that you were having that setup agent do are things that not necessarily admins don't like doing. It's like paper cuts." It's the little things that take 20, 30 minutes out of your day when you really wanted that solid hour to build that application kind of start to finish. Khushwant Singh: Indeed. Indeed. So we really viewed Headless 360 as the foundation upon which agents can work on behalf of the admins. And again, look, this is agents and admins working together because again, we've got to have the human presence to be able to validate, to be able to interrupt, to be able to initiate any actions that are performed by the agents. And if these agents can help drive greater productivity, then it's a win-win while at the same time having governance by humans over it. Mike Gerholdt: Right. Absolutely. Do you envision us being at a point where all of setting up Salesforce could just be a conversation with an agent based on the business and what they're trying to accomplish? Khushwant Singh: We are trying to do that. We are trying to do that. Now, I'll give you a view into our approach over here. We're taking a steel thread approach here where just trying to, for example, set up a customer service agent. That cuts across Service Cloud, cuts across platform, cuts across Agentforce, cuts across Data Cloud. So we've got to take a system view, an end-to-end view versus a feature view because that's what our admins go through. That's what a practitioner goes through. They look at it from a steel thread point of view, and that's what we as product teams have to do. And so as we think about using agents to go ahead and help provision, set up, get Salesforce up and running really quickly, that's the new strategy that we are taking over here, a steel thread view so that we can look across the board and say, "Look, which MCPs are missing? Which skills are missing? Which APIs are missing? Which metadata needs to be grounded? How do we ensure that across entire steel thread, quality is being ensured, governance is being woven into?" So we are starting that way and we want to scale that way as well. So my hope is that come Dreamforce this year, we'll be able to at least have a sizable number of steel threads that we feel really, really confident about on Agentforce really helping and agents helping in this particular side of things. So yeah, that's the approach we're taking, Mike. Mike Gerholdt: Wow, that's going to be amazing. Every time you think, "Oh, we've probably invented everything," technology just comes along and pulls the rug out from underneath of you and says, "Haha! There's new stuff. You got to figure this out now." I'd be curious, you're so ingrained in a lot of things that admins do on the platform, a lot of things developers do on the platform and then also having to keep up with AI. What is a piece of advice you could give people for how you personally keep up with the newest news and innovation on what's going on with AI? Khushwant Singh: Wow, that's a good question. I think first and foremost, I spend a lot of time just looking at reading publications like what's on TechCrunch, what's on Techmeme, trying to just understand how are technology providers adapting, how are companies using AI. That's one. The second one is actually you learn a lot via speaking with others in the community, right? Mike Gerholdt: Mm-hmm. Khushwant Singh: You talk to customers, you talk to other admins, you talk to other developers and you get to understand there are use cases that they're working through and some of the very many innovative workarounds or users of AI that they're applying. And then you're thinking about this that, "Wow. There's a trend here, there's something that we should be backing up and maybe providing out of the box within the product." And so I think that community piece speaking with other customers, with your peers, your colleagues from different companies, that sort of brings a very practitioners and practical view to things. So I think if you combine it with some of these ... So if you take a combination of both, what's being published out in the internet, what sort of bleeding edge and...
/episode/index/show/buttonclickadmin2/id/41287420
info_outline
Why Pattern Recognition Matters for Salesforce Admins
05/07/2026
Why Pattern Recognition Matters for Salesforce Admins
Today on the Salesforce Admins Podcast, we talk to Adam Stark, CRM Systems Administrator at Belmont University. Join us as we chat about how his experience as a musician with learning and pattern recognition has set him up for success as a Salesforce Admin. You should for the full episode, but here are a few takeaways from our conversation with Adam Stark. From the stage to the Salesforce Admin chair I met Adam Stark at TDX, and he had such an interesting path to becoming a Salesforce Admin that I had to bring him on the show. Based out of Nashville, he’s been a professional touring musician for twenty years. But when everything shut down during COVID, he spent his days on Zoom making an album, and his nights on Trailhead working towards his admin certification. Adam’s experience in the studio made it surprisingly easy to jump into automations on Salesforce—they made sense to him. “As a music producer, one of the things I got really good at doing was accomplishing signal flow, like trying to get a sound source to a final, presentable stage,” he says, “and that sort of signal flow process is the same with flows.” Whether it’s building tracks in a DAW or building solutions in Salesforce, Adam discovered that it’s still the same underlying logic. How pattern recognition makes learning Salesforce easier Starting out in Salesforce can feel overwhelming because the platform is robust. But, as Adam explains, the same could be said for learning guitar, and he realized that he could draw on his experience as a music teacher and performer. A part of learning any instrument is pattern recognition. You practice scales or licks in isolation so that it’s easy to find them and play them when you’re performing. “The more you do it, the more familiar you get, the more you begin to recognize patterns,” Adam says, “and once you see the patterns, things start to feel smaller.” Over time, something that seems very big, like learning a piece of music or trying to use campaigns in Salesforce, becomes more manageable. Acing your job interview with honesty I also wanted to hear how Adam got through the interview process and landed his first job as a Salesforce Admin. His experience as a musician helped here, too, because he was already used to doing interviews with radio stations while on tour. But nerves aside, Adam feels the key to his success was honesty. “I don’t know everything,” he says, “but if I don’t know it, I’ll figure it out, and we’ll find a solution.” For the folks out there who are still breaking into the ecosystem, Adam encourages you to get out there and meet working Salesforce professionals as soon as you can. Go to a community group, or even TDX, and pick someone’s brain. It can help you piece together what you’re learning in Trailhead by understanding what Salesforce looks like in action. There’s so much more great stuff from Adam about how he learned Salesforce and landed his first admin role, so make sure to listen to the full episode. And as always, make sure you’re subscribed to the Salesforce Admins Podcast, and we’ll see you next time. Podcast swag Learn more Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, Adam Stark didn't start in Salesforce. He started on a stage, building songs and touring as a professional musician. In this episode, he's going to share how that same mindset of thinking and systems, sequencing steps, and designing outcomes translated into building solutions with data, automation, and now AI. In this episode, we're going to dig into how admins aren't just learning tools anymore. They're designing how work actually gets done across people and technology. Adam's going to walk us through his journey from Trailhead to his very first admin role and what it really takes to connect those dots when you're building something bigger than just features. So if you've ever wondered how your past experience shapes the way you architect solutions today, this episode's for you. So with that, let's get Adam on the podcast. So Adam, welcome to the podcast. Adam Stark: Happy to be here, man. This is awesome. Mike: I'm glad we got connected through our architect friends. You are at TDX this year. So let's just give people an overview of who Adam Stark is and how you got started at Salesforce. Adam Stark: Yeah, absolutely. So I work down in Nashville, Tennessee. I work at Belmont University. And my journey into Salesforce is quite a unique journey. I actually was a professional musician for a long time. I was a touring artist, a professional songwriter, producer, and did that for many, many years. And then a thing called COVID hit in 2020 and really put a damper on the touring side of things, as you can imagine. So during COVID, I went through my own personal little existential crisis moment of, are we ever going to play shows ever again? This was a huge part of my income stream to provide for my family. And so I'm going through this crisis moment of what's next, what should I be doing? And a really good friend of mine who is a Salesforce developer, he told me, he's like, "Man, you should look into Salesforce. Just go on Trailhead and just poke around." And this is me not knowing anything about it. I didn't even know what a CRM was, but went on Trailhead and shout out to the Trailhead team because what an incredible resource, like unbelievable resource for people to learn the platform. Just really, really impressive. But yeah, I would spend my days working on music with my... I was a duo, if you will. I was in a band with another guy and he and I produced our records and we wrote all of our songs. And we're doing that virtually during the day during COVID where we're songwriting over Zoom, which is a complete interesting exercise to try to be creative over Zoom. And then we're recording audio files and sending them back and forth and compiling them into a record, which was wild. Doing that during the day. And then at night, I would just get on Trailhead and just start learning. And the more that I learned, the more interesting it got to me. And yeah, I got hooked and was really fascinated with what Salesforce as a platform was capable of doing and honestly how beneficial it is to businesses and organizations. Mike: Wow. Adam Stark: So yeah, I started learning. I was really determined to get an admin certification just to have in my back pocket. And then the world opened back up. And we did what artists called revenge touring because so many people were starved for entertainment at that point in time. It'd been a year, year and a half of no real public entertainment. So we got busy, went back out on the road, and I put Salesforce on the back burner for a little bit and then had more kids and talked to my wife and we just decided we're at a spot in life where being home is really, really nice. And I was pretty invested into the Salesforce journey at that point. So I went ahead and finished up, got my admin certification and I saw that Belmont was looking to implement Salesforce and went ahead and filled out an application and landed the job. Mike: So I've shown Salesforce to people. I've taken good friends of mine to user groups and I've had friends of mine even get up to Ranger status, which is a hundred badges and still be like, "Meh, it doesn't make sense to me or nothing's clicking." Something clicked for you that was like you got it and it's hard to describe, but I know what you're feeling, but what was it for you? Adam Stark: As a music producer, one of the things that I got really good at doing was accomplishing signal flow, like trying to get a sound source from one stage to a final stage, final presentable stage, if you will. And what I found was when I really started to get into some of the automation tools in Salesforce, especially flows, I realized this actually feels like a very familiar muscle to me. How do I get from A to B and what are the steps I need to take and in what order? What's the logic to get there? If I was producing music, it would be, I've got this... Say it's a vocal, do I want to EQ the vocal first or do I want to compress the vocal first? And then do I want to bus it to a different channel and affect it there with other things or do I want it to be isolated? And really that signal flow process is the same with flows. And then I just watched what it could do with data and I was like, "This is actually really fun because you're building solutions." As a songwriter, you're building songs. As a producer, you're building tracks and you're building records and I'm still building. It's just a different medium now. Mike: Yeah, that's fascinating because the next question I was going to ask you was... So I will confess, when COVID hit, I was one of those people that's like, "I'm going to learn a skill." Except I didn't do like you, I wasn't as devoted. I bought a guitar, a really nice acoustic guitar. And let me tell you why. Two reasons. One, I was like, if I'm going to learn, I'm not going to learn on some Walmart guitar level, I'm going to learn on something good. Adam Stark: Good for you. Mike: And two, if I buy something nice, then I feel like the investment will... I'll feel bad if I don't use the investment, right? Adam Stark: Mike, I'm the same way. When I told my wife, I'm going to get into cycling. It's like I need to buy a nice bike to make myself get on the bike. Mike: Right. I'm not going to buy the Lance Armstrong. I didn't go out and buy a Les Paul and spend a used car's money, but I got a nice acoustic guitar and people come over to my house, they see it on the stand. They're like, "Wow, do you play?" So here's the moral of the story. I took guitar lessons online for like a month solid and I realized, I don't know what it is, but music and math are two things that my brain's like, "Hey, that looks like I should go on the couch and sit and eat some Cheetos and somebody else will figure this out." That is exactly what my brain does. And so I would get to the point where I could learn the chords and figure out the finger and then they started to get to the point where they're like, "We're going to teach you music." And those notes on those bars, it might as well have been a foreign language. It was to me. It just didn't make sense. And I'm sitting here, during the day, I do Salesforce and we're dealing with complex flows and data validation and all this other stuff. And I'm like, "Why can't music make sense to me?" And it's funny because the reverse almost sometimes doesn't hold true. What works for you in understanding music did not work for me in taking understanding Salesforce to music. It's crazy. So anyway, so the moral of the story is I have a really nice acoustic guitar- Adam Stark: We're going to do some guitar lessons out there. Mike: ... that has sat on its stand. But I did play it for a month. Adam Stark: Well, you bring up an interesting point, and this might cross over to just some meta principles in general is the longer you're doing something, when you first get into anything, whether it's music or Salesforce for that matter, Salesforce is so daunting to get started because it's just so robust. There's so many things. It's overwhelming. It feels like this massive mountain you have to scale, but the more you do it, the longer you go about doing it, the more familiar you get, you begin to recognize patterns, right? And then once you see the patterns, things just start to feel smaller, if that makes sense. And I think the best way I could describe it is, that has been my journey with Salesforce. To get started, it felt like this massive undertaking. And the concept of campaigns, it's like, what are those for? I still don't really know why someone... There seems like there's several different angles to use campaigns, and it's just trying to get your head around it without ever being a user in Salesforce. But then you start seeing one example of campaigns, and then you see another, and then you're like, "Oh, well, now I've got some creative ideas because I recognize some patterns and I know what the functionality of this thing is." So yeah, I just think the longer you work at something, you recognize those patterns and then it's like, "All right, this actually is starting to make a little bit more sense to me." Mike: So you got your certification, you did Trailhead on and off when you weren't revenge touring, which is a fun term. I also remember that when the world opened back up and we were like, "Oh, let's do a thing." And maybe ironically, the first thing I did was I went to a Hyatt Regency and I got French fries. Adam Stark: Yes, sir. Mike: Totally boring, but man, the quality of French fries in those restaurant style fryers, nothing at home makes the same. Adam Stark: You are right. Mike: Everybody right now that's listening to this podcast wants me to ask this question, so I'm going to ask this question. Adam Stark: Okay. Mike: What was your first interview like? Because you obviously did a remarkable job of interviewing because you got a job. And I asked that question because I probably every day get a dozen or so DMs across various social networks of, I'm starting in Salesforce and I want to be an admin and how do I get my first job? Adam Stark: It is the million-dollar question, right? Mike: Right. Adam Stark: Again, going back to familiar muscles, one of the things that I was familiar with was the concept of interviewing because I did interviews with radio stations all over the country for years. So the idea of someone sitting in front of me and asking me questions, I wasn't expecting it. I was pretty nervous about it at first when I was venturing into this new career path because I was like, "I haven't had a job interview in how many years? It's been a long time." I was pretty nervous about it. And then once I sat down and started having the conversation, again, it was like a familiar muscle of like, "Oh, this is an interview just like a radio interview is an interview." So I felt pretty comfortable in the interview environment. And then it was me being honest. I really wasn't trying to put on that I knew more than I actually did. I wasn't trying to talk technical where I didn't understand it. I was honest with the idea of who I was as a person, what I'd learned. I'd achieved my certification. I have a pretty good base of knowledge for what Salesforce is and what it can do, but I don't know everything, but I'm determined to find out. If I don't know something, I'm the person that will figure it out. That's what I've always done. In my music career, where there were technical needs, we didn't always have resources to have somebody else meet those needs for. So I would be the one that would go and dive into it. I would figure out how to code lighting rigs and write DMX and program a light show for us because we needed a light show. So that's just my makeup is I'm a hard worker, I'm a learner, and I was honest about that in my interview. If I don't know it, I'll figure it out. We'll find a solution. And it worked to a point. Obviously, the organization took a chance on me because of an experience gap, but they saw something in me, I guess in my integrity and honesty, but also my drive to find solutions and not be afraid to go learn was attractive to them. Mike: Yeah. Well, from your perspective, what were you the most concerned about going into that interview? Adam Stark: I would say I was most anxious about the technical interview side of things, mainly because I have never been a front end user and I'd never been an admin before. So I was a little anxious about, am I going to look like a fool if they're asking me technical questions that I should know answers to? And there was a little bit of that. We found some limitations in that interview process, but again, I always said, "I'm happy to go learn. I'm happy to run that down and figure it out." Mike: Yeah. That's always, I feel like, the first question is, how do I get through the interview? How do I get through... I've never had a Salesforce admin job before. Aren't they going to look for experience? And to me, my first job, I didn't have a Salesforce admin certification or anything, and I just got handed that as a responsibility. I think it depends on the vision and commitment that people see. They saw something in you and said, "Hey, he can grow with us in the same way that we're probably looking to grow this platform." And also, you were a good fit culture. I think people are always like, "Well, what if I'm not smart enough?" Well, the other part of it too is, do they see themselves working with you? Adam Stark: Absolutely. That's a really good point. And I should mention, I don't think I mentioned this, I graduated from Belmont, so I went here for undergrad. Mike: A little bit of alumni. Adam Stark: I had the alumni angle coming in. So that really helped because I did understand the culture. I was here for four years. I'd know what the school represents. I knew how, at least on a basic level, how the school functioned and what some of the departments and teams were. So that was a starting point, right? But you're absolutely right. What you know is, I don't want to say half the battle, maybe more than half the battle, but there's a large percent of the battle is like, are you going to be someone we enjoy working with day in and out? And are you someone that thousands of other faculty, staff and students are going to be interacting with? Are you the person for that job? So you're right, the culture fit is a huge part of it and not to be underestimated. Mike: Yeah. What part, now that you're in it, of the job has probably been the most rewarding that you didn't anticipate? Adam Stark: Oh, it's for sure this idea of going through a journey with people. As admins, you find your users and you go through the discovery process with them and you learn about what are your pain points and you just learn about the language they speak and their staff, their workflows, all that stuff. And when you can deliver something that helps relieve those pain points for them, and at the end of it, they see that beautiful chart that they've been longing to see for two years, or they see the one-click automation that just executes perfectly and performs the need they need. When you can look at each other and high five at the end, and just it's so rewarding. I just love those wins with people. Mike: Yeah. The best part for me was always go live day. When somebody finally gets that solution and they're like, "Oh, I don't have to look at that green DOS screen anymore." God knows whatever else they were dealing with. Adam Stark: Absolutely. And then you see the spreadsheets and things that they've been managing. Mike: Oh, [inaudible 00:18:03]. Adam Stark: "Oh, Lord, this is your world." Mike: It is. But what's funny is some people, I'd say, almost get to the level of coding with the amount of cross tables and pivot tables and references that they build in these spreadsheets. You almost have a robot there. Adam Stark: You are right. You are correct. Mike: So you spent some time at TDX this year. Was that your first Salesforce event as an admin? Adam Stark: It was my second. I was able to go to Education Summit last year first, but I got to say TDX was awesome. I loved it. Mike: A lot of people that listen to the podcast don't get to go. Some people do the over under. I wish all my listeners could go to TDX, but for you, what were you going there thinking, "Here's what I need to leave with?" Adam Stark: It's a really good question. For me personally, I am not content staying where I'm at. I always have this need to grow. So as an admin, I feel like I'm getting comfortable with some of my admin tasks and I'm always...
/episode/index/show/buttonclickadmin2/id/41183915
info_outline
Agentforce Grid Enables Next-Gen Admins to Scale AI Workflows
04/30/2026
Agentforce Grid Enables Next-Gen Admins to Scale AI Workflows
Today on the Salesforce Admins Podcast, we talk to Avi Shah, Senior Director of Product Management for Salesforce AI. Join us as we chat about Agentforce Grid, a new way to coordinate data, automation, and AI agents. You should for the full episode, but here are a few takeaways from our conversation with Avi Shah. Spreadsheet-style AI workflows to simplify complex automation Salesforce Admins deal with data, automations, and AI every day. But how do you make everything work together in a way that makes sense? That’s why I sat down with Avi Shah to talk about Agentforce Grid, a spreadsheet-like tool for creating AI workflows. “Agentforce Grid is, in our opinion, the fastest and easiest way to build AI workflows,” Avi says. “You have columns for your data and the actions you want to take with it.” Some columns are AI-based, enabling you to run prompts or agents you’ve built, and others are action-based, allowing you to update records or call an invokable action to send an email. Put it all together, and you can build complex AI automations that can transform your organization’s workflows. Combining data and actions As Avi explains, Agentforce Grid gives you a simple, spreadsheet UI to perform powerful transformations on your data. You can pull things from Data Cloud, uploads, or even the web into a data column. Action columns give you a way to act. You can run prompt templates, agents you’ve already built, or inline prompts. Not everything needs to be an AI step, however—you can also perform more deterministic actions like formulas, updating records, or invoking flows. AI and non-AI actions work together in workflows All of this makes more sense when we talk about actual use cases. For example, you can use Agentforce Grid to assist with case categorization, working with a list of cases, a prompt column to analyze them, and another prompt column to look at those analyses and categorize them based on theme, priority, or issue. Avi has also seen customers take advantage of Agentforce Grid for transcript and session analysis for customer-facing agents. You can use the prompt column to analyze, classify, and extract information from transcripts to make sure that everything is working the way you want it to work. Be sure to listen to the full conversation for more from Avi on Agentforce Grid. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Blog Post: Admin Trailblazers Group Social Full show transcript
/episode/index/show/buttonclickadmin2/id/41061570
info_outline
How Agent Script Is Redefining the Admin Role
04/23/2026
How Agent Script Is Redefining the Admin Role
Today on the Salesforce Admins Podcast, we talk to Joshua Birk, Senior Director of Admin Evangelism at Salesforce. Join us as we chat about how Agent Script helps admins build more predictable and reliable AI solutions. You should for the full episode, but here are a few takeaways from our conversation with Joshua Birk. How Agent Script expands the admin toolkit The new Agentforce Builder gives admins new tools to create effective AI agents to help their orgs. Chief among them is Agent Script, which gives you more control than ever before over how your agents operate and behave. I sat down with Josh Birk to find out how it all works and why Agent Script redefines the admin role. Agent Script is a simple, high-level scripting language that lets you create complex instructions and actions for your agent to follow. You can step in as the human in the loop and create a framework for agent behavior and context, helping you build more predictable and reliable AI solutions. Redefining the admin role In simple terms, Agent Script lets you look under the hood at how your agents function. If you’re comfortable with coding, Script view allows you to make fast, precise changes and analyze error messages. Canvas view, meanwhile, uses visual blocks to make it easy to understand what’s going on with your agent. And of course, the Agentforce assistant is always available with suggestions if you get stuck. With so much more control over agentic behavior, Agent Script puts admins in the driver’s seat for how their organization will evolve with AI. You’re not just building a solution—you’re designing interactions between humans and agents. In the big picture, the role of the admin starts to look less like traditional system configuration and more like experience design, governance, and operational strategy. Getting started with Agent Script All of this requires a great deal of thought in terms of how you’re building new agentic solutions and rolling them out to your organization. Josh’s biggest piece of advice is to take it slow: “Don’t panic, learn Agent Script,” he says. Your existing agents will continue to work, so take your time to experiment with Agent Script to get a better understanding of what’s possible. Build some new agents and take it one step at a time. Listen to the full episode for more from Josh about what admins can do with Agent Script. And subscribe to the Salesforce Admins Podcast to catch us in your feed every Thursday. Podcast swag Learn more Salesforce Admins Blog: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: This week on the Salesforce Admins Podcast, if you've ever been wondering why you can't just copy and paste your old agents into the new version, this episode is for you. I'm joined by Josh Burke, who recently walked our admin relations team through what's really changing with Agent Builder and Agentforce. And more importantly, why it matters. We're going to unpack a little bit of hybrid reasoning, what determinism really means for admins, and how Agent Script gives you even more control without losing all of that magic. So with that, let's get Josh on the podcast. So Josh, welcome back to the podcast. Josh Birk: Thanks for having me, Mike. Mike Gerholdt: Well, we know that a few people are vibing now because of the dulcent tones of Josh Burke. So in all honesty, you were in a team meeting last week and you were kind of updating the team on stuff that's coming out for new products and beta products. And the biggest thing that you always explain is like, so why is this important? Josh Birk: Yeah. Mike Gerholdt: And of course, every now and then you end your little skit with, "Thanks for attending my TED Talk." Josh Birk: Yeah. Mike Gerholdt: And I remember last week's meeting, I was like, "Oh man, that's a podcast. I need to do that." Josh Birk: Yeah. Mike Gerholdt: Because the way that you would explained what was new with Agent Builder and Agentforce, I was just like, "Oh, that totally makes sense to me." So this is where we're at and this is what I was hoping we would talk about. Josh Birk: Yeah. And I'll have to try because I kind of remember getting into a very weird, almost like conspiracy theory level stream of consciousness. Mike Gerholdt: Oh, okay. Well, it didn't feel that way. Josh Birk: Good. I think that's years of experience of thinking in streams of consciousnesses, whatever the plural of that is. So yeah, so I think to kind of recap a little bit, so there were two things that I think are kind of important to consider about the new Agent Builder. One is what we mean by hybrid reasoning and what we mean by determinism. And then the second thing is, where does Agent Script fit into that? So the question we were kind of tackling was, why didn't we just go back and revise the old builder? Why can't we just reiterate on it? Why are we starting completely new? Mike Gerholdt: And I think you even said, not to interrupt you, but to interrupt you, why can't I just move my stuff over? Why do I have to rebuild it? Josh Birk: And why can't I just move my stuff over? Why isn't it just copy and paste? And Agent Script is part of that. So the original engine wasn't hybrid, is I think the easiest way to put it. So this came with a conversation with product management where I was trying to describe the differences between the old builder and the new builder. And they're like, well, let's ... The old builder wasn't doing anything wrong, so much is that the new builder's doing something more. And what the new builder's doing that is more is that when we say hybrid, when we think of AI, we think of AI being autonomous, making its own decisions. So we had the Atlas reasoning engine that's like, "Oh, Mike is looking to find a description on a certain account. I'm going to take these actions. I'm going to put them together into a plan and then I'm going to try to make that work." What we didn't have was the human in the loop portion of that. And where Agent Script really comes into play is it allows users to basically be like, "In this scenario, I want you to behave like this." So when we say hybrid, we mean, yeah, we're letting the engine kind of do its own magic. But then we're also applying almost ... I don't want to say flow because that's such a loaded term for a Salesforce community. Mike Gerholdt: Yeah. Josh Birk: But well, let's go to the name, right? A script, right? A script that a human can actually control personality, control tone, control behavior. When you see this, do this, when you need to call it human. So kind of giving this very human-friendly dialogue that you can then apply to the agent. And so that's how we provide more determinism. So the reason you can't just copy and paste your old agent into the new agent is the old agent doesn't have that connection to Agent Script. And so it's literally missing a piece of the puzzle to run the engine kind of thing. Mike Gerholdt: Right. No, I think ... I mean, hearing you explain it, sometimes, at least as an admin, I was always like, "Oh, so they just came out with another version because they had to stick another layer of something on." Josh Birk: Yeah. Mike Gerholdt: Or like a car analogy, they changed the fuzzy dice. But this is, no, no, no, no. We're really moving from this version, which ran a certain way to an even better version. And that's the part that I feel like ... Because you brought up Flow, I feel like we've been through this with Flow. Josh Birk: Yeah, totally. Mike Gerholdt: We got like a business process ... Business management process is what it used to be called. And then that thing went through iterations and then we found a different engine to run it on. And then we had ProcessBuilder for a while and then now we have the new Flow. And it's like sometimes it's not just the interface that changed, but it's actually the guts behind it. Because I mean, I can't recall ... It's sometime between after the pandemic and today that AI came around. And I feel like that's the window that works in my brain because I can't put down a time. Josh Birk: Totally. Mike Gerholdt: Was it 2022 or is it 2023? I don't know. Post pandemic, before today, AI happened. But just when you thought, okay, I've kind of got this thing figured out, nope, all of this stuff changes. Josh Birk: Yeah. And it happens in technology, right? It happened with the Lightning Web Components. We really kind of had to go back and take the new ECMAScript format and take the new ECMAScript standard and really rebuild the way our web components work so that they work in a standard based way. So that they're more similar to React and more similar to these other frameworks. And there wasn't really a way to go back to Aura, now it's called, and just be like, "Hey, Aura, be more standards based." Mike Gerholdt: Right. Josh Birk: Because Aura was effectively just ... It was pre-standards is the easiest way to put it. So this happens, right? Now, if people are listening to this and like, "Oh gosh, I have all this work in my old Agent Builder, what am I going to do?" First of all, Salesforce does not like taking away people's toys. Mike Gerholdt: Right. We've learned that. Josh Birk: We've learned that. I've said this anecdote in workshops so many times because when we did release Lightning, one of the most common questions I got was, "Are you going to end of life Visualforce?" So I went to the product manager, might have been Skip at the time. And I'm like, "Hey, when would you end of life Visualforce?" And his response was, "When I pull up a dashboard and I see there are zero active users using Visualforce, and then I'd probably wait three months after that." Mike Gerholdt: Just in case. Josh Birk: Just in case, right? Mike Gerholdt: Yeah. Josh Birk: So the classic one is not going anywhere anytime soon. But I think the other important thing is a lot of our messaging when it comes to building out agents is that you're already an Agent Builder. You're already somebody who can build AIs because you know Flow, because you know Salesforce. Now you've learned topics, now you've learned actions and all of that kind of stuff. All of that is still true. It's all still true. And so you're still getting to use your existing skills in order to control it. It's really just adding the skill of knowing how to write out the Agent Script. Now, one of the cool, really cool things about ... Well, there's two cool things about the new builder that I think will really help people with that. One is the Canvas mode. And so Canvas mode takes what you write in Agent Script and puts it into a visual UI interpretation. And so this is very handy for two things. Number one, it gives you a little bit of kind of a predictive window as to what your script is going to be doing. It also gives you a very visual way of thinking about it. It's also a really neat way to go look at somebody else's Agent Script and translate it. So you don't have to read through all the thing. You can flip over to Canvas mode and be like, "Oh, that's trying to get these parameters in order to do these actions." The other thing going back to vibes and this new AI coding AI is that you'll have an AI assistant in Builder to help you build the AI. So you're not alone. Agent Builder itself will also help you create these things out. So my biggest statement back in the Lightning web component today was don't panic and learn lighting. And now it's, don't panic and learn Agent Script and you're going to build out some really powerful solutions. Mike Gerholdt: Well, and I think back to ... I mean, I was chatting with Jennifer today too, and just the way that we started using AI. And I remember, I mean, not that long ago doing workshops and saying, "Okay, you need to tell the AI your role, its role. And then what you want it to do and what it can do and what it can't do." And for everything, whatever model you were using, I remember writing prompts a certain way. And I feel like for this and with Agentforce, especially for admins, this is the least set it and forget it kind of thing. Josh Birk: Yeah. Mike Gerholdt: Because you can always go back and make it better. Josh Birk: And tweak it. Yeah. Mike Gerholdt: And this is- Josh Birk: And tweak it in theory. Yeah, exactly. In very small corners too. Mike Gerholdt: Small corners that make a huge difference because like with Flow, I mean, half the time I'm just excited that I got it to do what I needed it to do. And then I look at what Jennifer built and I feel like I'm over here rubbing two sticks together. Josh Birk: Together. Mike Gerholdt: But with Flow, that's your goal, right? You just want that, okay, boom, it's done. But with agents, people are interacting with them, they're asking them questions, I can make this so much better and I can make this better. Here's the clue when people stop using them. To your point of when you would replace Visualforce. Well, if you're looking at your Agentforce dashboard and seeing, "Hey, your users aren't using it." well, now might be a time to like, "Well, let's switch over, let's round some edges, let's change some things." And I would say this is like an every two weeks kind of thing. Which sounds crazy, but the amount at which you're going to find feedback from your users. And the amount of rounding of edges, I worked myself into an English problem here. It's going to be like, you're going to find like a star and then it's going to look like a ribbon of rounded edges that you have to work on. Because as you use that new builder, now your users are happier and being more proficient. And I think like even today, when Jennifer and I were talking, we were talking about building Gems in Gemini. And I said, "One thing that I do is I'll actually have another window open and ask it to refine my instructions in Gemini." Josh Birk: Oh, yeah. Mike Gerholdt: Because the way that I write the instructions works for me, but maybe not for it and it can refine it. And I've seen it take thousand word prompts and reduce it down to 200 and it's actually faster. And so having all of that is sometimes there's a lot to keep up with. But it's all for the better because now you're like click click, oh wow. And then you get that ... I think it's always, at the end of the day, the admin is going after that first look on a user's face and they're like, "Ooh, this- Josh Birk: Nailed it. Mike Gerholdt: ... just made my life so much easier." Josh Birk: Exactly. Yeah. And I think to kind of append to that, again, not to overuse this analogy of Visualforce. But it's like don't be in a rush to go and recreate your existing agents in the new Agent Builder. If you have a new agent you're building, I would start with the new Agent Builder and go from there. Two reasons, don't fix what's not broken. And also you'll learn more about Agent Script and things like that. So when you do decide to move your old agent to the new builder, you'll be more educated on how to do it. Now, huge forward-looking statement, asterisk, safe harbor, there is a conversation internally about creating a porting tool of some kind. It's on a roadmap somewhere, no guarantees. Mike Gerholdt: Oh. Josh Birk: But I do think people should know where we are looking at that kind of thing. So stay tuned because we do want to make this as easy as possible. Mike Gerholdt: Yeah. Yeah. But also there's some joy in rebuilding things and maybe that's just me. But I've always- Josh Birk: Seems like part of the tech world. Yeah, exactly. Mike Gerholdt: Yeah. I mean, there's that and there's also, it reintroduces you to things that at one time you were good at and brings it back. And then I think too, anytime that these new features come out, one part is, okay, well, here's the tech to change it. But also, has the process changed? Has the stuff that people do internally changed? And that might also be true. And the other point you brought up, which was really salient, just because you can change it, doesn't mean you should right now. Josh Birk: Yeah. Mike Gerholdt: One thing that ... And I was a part of this, doing a big org rollout. What are the other changes that your users are experiencing within your organization? And so you adding to that, is that one layer too many? Or when would be a better time? Like we always joke, like you're not going to roll out a sales opportunity change the last week of the quarter. Josh Birk: Right, right. Mike Gerholdt: So picking your timing, this means that gives you time to test. Josh Birk: Yes. Yeah. And I think this goes into a lot of the conversation we're having about admins in the AgentiCare. Like what kind of a steward are you being? And how are you using those existing instincts and skills as an admin to not topple over the apple cart, but now you're doing with AI? Mike Gerholdt: Right. Because in my opinion, admins are getting paid for their judgments. Josh Birk: Yeah. Yeah. Agreed. Mike Gerholdt: Josh, thanks for coming on and helping us understand the new Agent Builder. Josh Birk: Thanks for listening to my TED Talk. Yeah. Mike Gerholdt: I knew you had to fit that in. And we've got just enough time that Daryl can hopefully get his boat from home to the internal ... It's not internal. Inside the podcast joke. Josh Birk: Inside the podcast joke. Mike Gerholdt: You have to go back to the Daryl Moon episode- Josh Birk: Yeah, exactly. Mike Gerholdt: ... and then it all makes sense. It's the key to unlocking the mystery of the Dan Brown episodes. I don't think so. Josh Birk: I love it. Mike Gerholdt: Anyway. Josh Birk: I love it. Mike Gerholdt: Have a good day, Josh. Thanks for coming on. Josh Birk: Cheers, my man. You too. Mike Gerholdt: Huge thanks to Josh for coming on very impromptu and turning what could have been, I think, a little confusing of an update into something that actually makes sense. New engine, new script layer, AI helping you build AI. Awesome. I like it. So remember, don't panic, learn Agent Script, and it's all about having fun. And like any powerful tool in Salesforce, it's less about flipping a switch and more about steady improvement over time. So if this episode helped you clarify your next steps, share it with another Salesforce admin who's navigating some AI right now. And until next time, we'll see you in the cloud.
/episode/index/show/buttonclickadmin2/id/40979540
info_outline
How Salesforce Built a Scalable AI Puzzle App in Six Weeks
04/16/2026
How Salesforce Built a Scalable AI Puzzle App in Six Weeks
Today on the Salesforce Admins Podcast, we talk to Jagan Nathan, Senior AI Architect at Salesforce. Join us as we chat about how he built a Slack app for Salesforce’s “Million Dollar Puzzle” ad, where millions of concurrent users raced to solve riddles by chatting with Slackbots. You should for the full episode, but here are a few takeaways from our conversation with Jagan Nathan. The tech behind the “Million Dollar Puzzle” contest If you were watching the big game, you might have seen Salesforce’s “Million Dollar Puzzle”. As soon as the ad aired, the race was on for contestants to solve puzzles by chatting with Slackbots to discover the location of a secret vault. My guest this week, Jagan Nathan, built the Agentforce-powered Slack app that made everything possible. They needed Slackbots that could act as a conversational gateway for millions of concurrent users. And with a million dollars on the line, they needed to be sure it couldn’t be tricked into giving away the answer. Even more incredibly, Jagan pulled all of this off in only six weeks. I sat down with him to find out how he did it and what he learned along the way. Using AI to build quickly at scale Just like with any project, the first step was to gather requirements—though it’s more fun when your stakeholder is the “Puzzle Master.” Jagan and his team needed a thorough understanding of what they were dealing with and how they could translate those gaming mechanics into the application. The clock was ticking, and AI was pivotal for accelerating the development timeline. Jagan and his team used Salesforce Vibes for quick prototypes and wireframes to help them decide what to build. But the development team still needed to take these ideas the rest of the way. “AI cannot help us solve all of the architecture problems,” Jagan explains, “we have to have a human in the loop.” With millions of concurrent users, they needed to do thorough testing—and quickly. Luckily, they could use Salesforce Scale Center to stress-test for performance issues and identify bottlenecks so they could be sure that when the ad ran, the app would work flawlessly. Designing for security with a million dollars on the line I know this might be shocking, but when you’re holding a contest with a million-dollar prize, some people will try to cheat. So Jagan and his team needed to make sure there were plenty of guardrails in place to make sure someone could win it fair and square. They needed to start with a security mindset and think through any security vulnerabilities as they designed the application. Einstein Trust Layer was crucial for toxicity detection and monitoring for anything malicious, like prompt injections. Most importantly, they were very careful with what information the Slackbot knew and what it didn’t. It didn’t have the solutions to the puzzles, so even if someone managed to crack it, they wouldn’t be able to get very far. Instead, puzzle answers were always validated by humans. There’s a lot more from Jagan about building for scale quickly, so make sure to take a listen. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Blog Post: True to the Core Deep Dive: Salesforce Admins Blog Post: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: When you think about building on Salesforce, it's easy to focus on features. But today's admins aren't just implementing features. They're designing systems that have to perform, scale, and hold up under real world pressure. In today's episode, I'm going to sit down with Jagan, an architect who helped deliver a high stakes, high scale experience serving millions of users through a custom Slack-based AI system. We're going to unpack what it actually takes to compress months of work into weeks and how to design for trust and guardrails from day one, and where human judgment still matters in an AI-driven architecture. So if you've ever wondered how your role evolves from building flows to orchestrating full systems, this is the conversation because this just isn't about speed. It's about responsibility at scale. Now if you enjoy this episode, hit Subscribe, share it with a fellow admin and let's get Jagan on the podcast. So Jagan, welcome to the podcast. Jagan Nathan: Hi, Mike. Thank you for having me. Mike Gerholdt: Yeah. Well, I was excited when you sent me the Slack DM about some of the projects that you've worked on here at Salesforce. But for people that don't know you or maybe didn't listen to the episode of the Developer Podcast you were on, could you tell us a little bit about what you do at Salesforce and the fun project you got to work on recently? Jagan Nathan: Sure. Hey everyone. I'm Jaganir. I'm architect part of our AA practice team here at Salesforce, building agents for our customers. This is my third podcast on our channel. My previous episodes were focused on event monitoring and threat detection. Mike Gerholdt: Wow. Let's talk about some of the stuff that you got to work on around the big football game that happens in February that Salesforce had an ad on. Jagan Nathan: Yeah, sure. Very excited. This is our first ever online puzzle app, which was built entirely on our Salesforce platform. Mike Gerholdt: So you got to be a part of that. I'd love to know from a Salesforce admin listening to this, we used the platform for the puzzle. Can you give me a little insight into that? Jagan Nathan: Yeah, so let's unpack this. So what we did is we didn't just build a chatbot. We deployed our custom Slackbot, of course, powered by our Agentforce, which acted as a conversation gate for millions of participants grounded in real-time data. So we just collapsed a nine-month roadmap into just six weeks, which is 42 days. Within that, we designed iterator and we went live on day of February 8th. Mike Gerholdt: Wow. Nine-month roadmap. Holy cow. That's insane. I mean, can you talk a little bit about the architecture problem solving and how you moved so fast so quickly? Jagan Nathan: Yeah, sure. Of course, we are a customer zero company. So we leverage our internal tools and the external tools, what we have developed. So for this entire application, we use our Agentforce Vibes to accelerate the development. We quickly do a lot of wireframes and the prototypes, and then we figure out what sort of user experience and what sort of applications we need to build using the power of Agentforce Vibes that accelerated our development timeframe to start with. Mike Gerholdt: Oh, that's nice. I mean, I've definitely seen Vibes do a lot of things. I'm assuming we really took it out for a test spin on this one. Jagan Nathan: Yeah. So we use Agentforce Vibes to do this. And then we also leverage our scale testing product because this particular application, we anticipated there's going to be at least millions of users who is going to access this application concurrently. So performance is super important for this application. So we leverage scale testing product to identify quickly and resolve all the problem bottlenecks as well. Mike Gerholdt: Yeah. Wow. I've had some talks with PM for, was it Scale Center? Is that what you used? Jagan Nathan: Yep. Yes. Mike Gerholdt: Yeah. And we've put out a couple of blog posts about those. So for admins that aren't familiar with what Scale Center is and what some scale testing helps them do, what did you use it for? Jagan Nathan: Yeah, Scale Test is the performance evaluation tool. So that gives a lot of metrics. Before customers used to run some automated performance testing to figure out what sort of performance bottlenecks they have in the ACRM instance. But with the power of Scale Test, we can write a test plan creation. We can do a test environment set up. And we can also simulate all those real-time performance challenges, and then we can schedule those test trend to see where exactly fails and at what threshold it fails so that we can understand the spike and then we can fine tune our application. Mike Gerholdt: Okay. That's super helpful. I mean, was there any surprises that happened in your testing? Jagan Nathan: Oh, yeah. Since we anticipated there's going to be at least millions of concurrent users, there's going to be a lot of DB rights back and forth. And then we ran into a lot of issues. We were able to figure out with the help of scale testing, we figured out all those in the sandbox, and then we fine-tuned our code and the logic and everything to make sure it is scalable for concurrent users as well. Mike Gerholdt: That sounds like a lot. Holy cow. I don't think I've ever built an application where I had to plan on a million people using it concurrently. So you talked about Agentforce Vibes. I'm sure we leaned into a lot of AI to build this. How did you decide really what AI should do versus what humans should still control during the project? Jagan Nathan: Yeah, that's a great question. So of course we can use AI, but AI cannot help us solve all the architecture problems. One of the architecture problems, we have to use human in the loop. For an example, when we started designing this application, considering the scale of users, we thought just by writing a prompt, it's not going to solve the problems. Prompt is going to give, based on the request prompt is giving the response back, but that alone is not sufficient enough. So that's where we use humans power to understand and architect well for this [inaudible 00:06:50] game. Mike Gerholdt: Yeah. Wow. On top of that, you got to think of guardrails because I'm sure somebody was thinking, "Well, since Slackbot was in use, what if somebody just asked it, what's the answer?" That seems like the most obvious guardrail to put in place, but how did you sit down and think about some of the guardrails to put in place for an experience like that? Jagan Nathan: Yeah. So there's an aha moment here. So when we built this custom Slackbot, we intentionally built in a such a way that this custom Slackbot doesn't know the puzzle answers. So even if someone tries to crack the puzzle Slackbot, they would not be able to get the answers out of it. So that is the first so that no one walks away with the puzzle answer, right? Mike Gerholdt: Right. Jagan Nathan: So that is the first thing what we did. And then the next thing what we did is automated toxicity detection at the hedge. So what we tried to do is we leveraged, of course, the power of Einstein Trust Layer and then it's not like a traditional AI chatbot. So we just built intentionally a custom Slackbot with all those automated toxicity detection so that if the Slackbot finds some harmful response or some sort of response which is not supposed to do it, if the user on the other side, if the user is a bad actor trying to do some prompt injections or whatnot, our platform is able to monitor and mitigate the risk immediately. It's being very proactive stage of monitoring. Mike Gerholdt: Yeah. No, I guess I just thought the Slackbot would have access to the answer and you'd just have to guard it really well. But if you set it up and never give it access the answer, then it can't accidentally hallucinate and give the answer. So wow, that was probably a little bit of planning on your end. So looking at what you built and deployed, if a Salesforce admin wanted to recreate some of the best learnings that you had from this, what were some of the key takeaways that you had? Jagan Nathan: Yeah. So I had that security mindset to start with. Let's say you are a bad actor or else you are trying to act the system, what sort of loopholes you might find in the system and then find all those security vulnerabilities, why you designed the application itself. It's like having a security mindset is super important. And that too for the scale of the cash prize is one million. So definitely there's going to be a lot of participants going to crack the Slackbot or else they try to trick the Slackbot and try to get the puzzle's data out of it. So I would say security mindset is super important. And then start with the understanding of, basic understanding of how prompts work and then try to build a Hello World app in Slack to start with. Mike Gerholdt: Yeah. Did you have a whole team of people that were just, I guess it's called negatively testing because there's always positive testing, which is test it to make sure it does what it's supposed to, but then there's negative testing, which is to make sure it doesn't do what it's not supposed to, right? Jagan Nathan: Yes. So we did advisory testing as well. So of course, we used the power of AI. We built our own agents to try to craft this agent to see where it bakes. And then we try to enhance our prompts and all the safety guardrails to make sure. We did a ton of testing during our design and implementation phase just to make sure Slackbot is not supposed to provide any information which is not needed or which is kind of out of privacy zone. Mike Gerholdt: Yeah. Wow. So given the scale of everything that you did, what is something that you maybe hit a limit with AI that you didn't know would exist? Jagan Nathan: What we felt down the road is like, AI cannot solve all the problems for sure. That too for the scale of millions of concurrent users trying to access the system. We thought there might be places in which there might be some timeouts happening, but knock on the wood, nothing happened. We were able to build a scalable application. Now, we don't rely 100% on AI. We also added a human in the loop, for example, some sort of puzzle answer validation, mainly the final puzzle answer validation. We don't just rely only on the AI. We use a human in the loop. So that we brought in experts on the puzzles who validated the answers and lot of stuff at the backend. Mike Gerholdt: Yeah, that's interesting. What was it like to have to translate a puzzle into a technical requirement? Because I feel like a lot of Salesforce admins often get business processes that can sound like a puzzle and they have to translate that into a technical requirement. Jagan Nathan: Yeah. Great question. That was very interesting and challenging because not all of our teammates worked on the puzzles or solved the puzzles before. If you take me as an example, right? I haven't solved any puzzle in the past. I went through the basics of understanding what sort of puzzles exist in the real world. Some are logical puzzles, a lot more Sudoku puzzles, some are mathematical puzzles, some are geolocation puzzles. So for us, it took some time for us to understand the basics of what sort of puzzles we have. And then we had conversations with the puzzle masters and then who built this puzzle. We tried to understand from the puzzle master what sort of application they wanted to build because they know very much on the gaming side and we know very much from the technical side. It goes hand in hand. That was very interesting and challenging conversations we had to understand. It took some time for us to understand and translate the gaming mechanics into our application. Mike Gerholdt: Wow. Okay. So be honest, how cool wouldn't it be to have a title named Puzzle Master? Right? Jagan Nathan: Yes. Mike Gerholdt: So one of the last questions, because I bet you use Slack a lot as the project team to get things done. I've noticed as I work at Salesforce and we use Slack in different channels, I've noticed changes in behavior, ways that we work faster with Slack. Was there something you left this project with where you're like, "Oh, I'm going to start posting this way in Slack or I'm going to start sharing information this way in Slack," because it was something that actually was a way that you worked in that team. For example, I've noticed that when we have really long threads rather than posting one big post that could take up a lot of screen, somebody will post a comment thread about a topic and then put all of that into the comments because then it keeps their feed kind of cleaner. Was there anything like that that you kind of learned as like, "Ooh, I'm going to start doing this more often now in Slack?" Jagan Nathan: Well, yeah. So what we did is when we started working on this project, we had a 14 member team, few members were focusing on the technical side, few members were having conversations with the puzzle master. So we used Slack as a project management tool. So what we did is we created a Slack Canvas with a list of team members and then what sort of priority task we need to work on. And then we automated to the next level by adding a Slack list as well. We created a list, like what is the roles and responsibilities for each team member? And we started tracking the list. And then every week we were running the Slack workflows. We extensively used the Slack workflows. We created a ton of workflows, one workflows for summarizing the Slack channel so that what happened in this one week's ... Because see, we had at least six weeks to develop and build and deploy, right? Mike Gerholdt: Yeah. Jagan Nathan: So each week we were running the Slack summary workflows to understand what is the latest going on around them. And then we were extensively using Slack, mainly the Slack Canvas and then the Slack list to start with. Mike Gerholdt: Yeah. Wow. I love building workflows. I tell you, they're really cool. I can almost build too many sometimes. Well, I appreciate you coming on and sharing what you could about this. I think the amount of work that was completed for this project and in the amount of time is super incredible and I'm sure it was quite the learning for you. So it also teaches us that even the biggest projects we can still accomplish very quickly. So thanks for coming on the podcast. Jagan Nathan: Thank you for having me, Mike. Mike Gerholdt: So it was great of Jagan to come on and share his experience on that project. I think it's a good reminder that building on Salesforce today means thinking beyond automation. It's about really designing systems that are resilient, trustworthy, and ready for real world scale. From AI guardrails to perform testing to human in the loop decision-making, this is the work that the modern admin is balancing speed with responsibility. Now, if you found this episode helpful, share it with another Salesforce admin who's thinking bigger about the systems they run and making sure that you're subscribed so you don't miss what's next. Until next time, we'll see you in the cloud.
/episode/index/show/buttonclickadmin2/id/40871185
info_outline
How Admins Can Get the Most Out of TDX 2026
04/09/2026
How Admins Can Get the Most Out of TDX 2026
Today on the Salesforce Admins Podcast, we talk to Jennifer Lee, Lead Admin Evangelist, and Eliza Riley, Content Manager for Admin Marketing at Salesforce. Join us as we chat about what’s coming at TDX and how to get the most out of your visit. You should for the full episode, but here are a few takeaways from our conversation with Jennifer Lee and Eliza Riley. Connect with product leaders at TDX TDX is around the corner, and I sat down with Eliza and Jen to talk about must-see content for admins and how to get the most out of your visit. One thing you won’t want to miss out on this year is Sneak Peeks, a special experience in the theater where you’ll be able to see new features in action. We’ll be showcasing things like the new App Studio and Setup with Agentforce so you can get an idea of what’s coming. There will also be a Q&A with the product teams so you can give feedback and better understand the roadmap. There will also be True to the Core Deep Dives covering flow and automation, reporting and analytics, Agentforce, and low-code and pro-code development. These extended Q&As offer you a chance to connect with product leaders and ask all of your burning questions. Finally, if you have a thorny problem that needs solving, you can sign up for a 1-on-1 Ask the Expert session with a Salesforce engineer or product leader. See the sights We’ve been talking a lot on the pod about Agentforce Vibes, so maybe it’s time to see it for yourself. In the Agentforce Vibes Zone, you’ll be able to get hands-on with how you can build using natural language. There will be demos, a guided project, and sessions to help you get started. You’ll also want to check out the Hackathon Showcase in the keynote room, where the three finalists will pitch their builds live to the judges. As Jen says, it’s always interesting to see how different people solve problems and get some inspiration for what you’re building in your org. If you’re coming alone, make sure to stop by the Community Cove. It’s the spot for networking with other admins and other folks in the ecosystem, and a great jumping-off point for exploring TDX. How to plan for TDX As you’re getting ready for TDX, it’s important to remember to balance planning with flexibility. “You're going to be connecting with people,” Eliza says, "you might see a cool activation that catches your eye, or a robot you might want to stop and grab a picture with.” Wear comfortable shoes, bring water and snacks, remember that some sessions require headphones, and make sure you have a swag bag. Most importantly, come find us and say hi. Make sure to listen to the full episode for more on what’s coming at TDX. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce Admins Blog Post: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Love our podcasts? Subscribe today or ! Full show transcript Mike: If you're heading into TDX thinking it's just another event, this episode might change how you show up entirely. Today, I'm joined by Jennifer and Eliza to break down how to actually get the most out of TDX26, not just attend it. We're going to talk about everything from Sneak Peeks and Ask the Expert zones to hands-on Agentforce experiences and so much more. Most importantly, how these moments are going to help you design better systems, not just learn new features, because this isn't about collecting swag or just session notes. It's about connecting with the people building the platform and shaping how your organization runs. So, if you're serious about evolving from admin to orchestrator, we got your game plan. So, let's get Jen and Eliza on the podcast. So, Jen, Eliza, welcome to the podcast. Jennifer: Thanks for having us. Eliza: Super happy to be here. Mike: It's going to be fun. So, Jen, you're a veteran of the podcast and in the community. Tell us a little bit about what you do at Salesforce, and then, we're going to get into what we do for TDX. Jennifer: Sure. I'm one of a few admin evangelists on the team. So, some of the things that I do on a daily basis is write blogs, create video content, and present to all of you admins out there. Mike: Just a few. Eliza, you're like our secret Alfred in the bat cave. I feel like all of us evangelists come back with torn suits, and our Batmobiles have a flat tire and you fix everything up. But you also did a whole bunch of stuff at TDX last year, but what do you do on the admin team? Eliza: Yeah, I know you guys are really on the front lines. So, I'm a content manager on admin relations, and I focus on creating resources and storytelling to help admins grow into more strategic roles. And I also lead our admin blog. So, really helping to translate just everything that's happening across Salesforce into what actually matters day-to-day. And that includes a lot of coverage on TDX, which I'm really excited about. Mike: Yeah, we have a lot planned. To be fair, you also did a showdown last year at TDX, if I recall. Isn't that right, Eliza? Eliza: Yes, we did. That was super fun. We had a few members of the community come and showcase their best Agentforce use cases. We had some prizes, and it was just a really great time. Mike: Right. It was. You had one vivacious MC. Should hire him back. Let's talk about this year though, because Eliza, you wrote a blog post to help admins get ready and write down their list of stuff they can't miss for TDX. And the first thing is the Sneak Peeks area, which you and Jennifer are working on. So, let's hear a little bit about that first. Eliza: Yeah. So, Sneak Peeks is this really special interactive theater experience. And these are really building on the success that we saw at Dreamforce. So, if any of you attended Dreamforce last year, Jen did an incredible job in the Admin Meadow creating this special exclusive look at what's coming next for Salesforce products. So, we're building on the success of that this year, and we're going to be showing you quick demos of all of our in-progress innovations, and then, opening it up for live Q&A with the product teams. And so, what's really cool about this for those of you attending in person is that you might be some of the first people to actually see these developments in action. And then, if you're not able to join us and you're watching from home or tuning in later, we'll be covering some of these on the admin blog as well. So, you definitely won't be missing out on any of the fun. Mike: Oh, nice. Jen, you got to peel back a layer. Tell me, give me something. Give me a sneak at the Sneak Peek. Jennifer: Well, first I want to say that as a former customer, when I went to TDX, one of the things that I look forward to was seeing the roadmap and interacting with the PMs. So, I'm really excited to be able to bring this to all our technical audiences. But a couple of things that you will be able to peek into, we'll be introducing the new App Studio. There's also Salesforce Release Manager. So, think of this as being able to say, for these releases, "I want to enable these things in my org." That's the idea behind that. And then, we're just bringing on a lot of... We have three sessions on flow, like thinking about new flow design or what's coming up with screen flows or how to manage your flows. We're also bringing in set up with Agentforce, not only in Salesforce itself, but also in Slack. So, there's just a lot of cool things, but I'm just going to tease a few. Mike: Yeah. No, thanks. You got to show up, but that's why it's called a sneak peek, right? I'm excited for this. I agree. When I was a customer, being able to just see what's new, take those notes, most of the time feel, like the plane ride home, you could be like, "Oh, cool. I don't have to build this because it's already going to come as a feature," is something that I always look forward to. Now, you mentioned we could do some Q&A with the product teams at your Sneak Peeks, but we also have something called Ask the Expert Zone. Can we talk a little bit about that? Jennifer: Yeah. So, we had that also in the Admin Meadow at Dreamforce, and it was a huge success. So, think of it as you come to TDX and you have this burning question in the back of your head or this problem that you're trying to solve for and you're just hitting a wall, you can come and talk to our experts. And the difference between what we did at Dreamforce and TDX is now, we have specific topics. So, instead of going to, let's say a generalist that's supposed to know all the platform, and who possibly can because Salesforce is so huge? There's no way. You're going to go and talk to a specialist. You have that burning of question about automation. You're going to talk to the flow expert. Or you need help with getting started with Agentforce. You're going to talk to the Agentforce expert. So, we have these various topics, six different topics, and you come during that scheduled time and you'll be able to have one-on-one discussions with those experts. Mike: I like that because we did a version of that, and I volunteered for it, and I think it was just about every other person, I could answer questions. Jennifer: Yeah, it's hard. Mike: I know. Man, I just felt bad, and I was like, "Well, I can't really answer your question, but do you want to work through the problem together and maybe we'll stumble across something?" One of the things that I think is important and along those same lines, at a surface level, it's like, "Oh, cool. So, I'm going to be able to talk to Data 360, or flow, or user access and governance." I see those as topics. Waiting until you're at the event, Jennifer, is not the best idea. If you were a customer today, you have 14 days to Dreamforce, oh my goodness, to TDX, what would you do as a customer to get ready for those Ask the Expert sessions? Jennifer: I would hone in on what are the top things that I want to know from that expert? Come up with some questions to bring to the table versus just showing up and coming up with questions off the top of your head. Or if you have a problem with the solution, bring your laptop and show the person, "Here's my problem. Can you show me where I'm going wrong?" That way, you're not generalizing it, and then, they could really help you because they're actually seeing the thing versus saying like, "Oh, well, I think it might be this," or, "Maybe it's this, but I don't know because I don't know the details of your problem." Mike: That's really good. And when we did that version of Ask the Experts, I had somebody print off their entire org schema and bring it. Jennifer: Wow. Eliza: Oh, wow. Mike: So, yeah, it was awesome and also kind of sad because I was like, "Man, it's a whole forest of trees you just killed, but I know exactly how your org's set up." It's wonderful. Speaking of experts, we have something new, and I know Kate's been on to talk about it, but Eliza, True to the Core Deep Dive, that sounds new. What's that? Eliza: Yes. This is one of our new activations or tracks that we're really excited about this year. So, True to the Core Deep Dive is essentially just an extension of the larger True to the Core keynote that we all know and love. So, it's going to give everybody the chance to just go a little bit deeper into specific topics. So, they're essentially extended breakout sessions letting you connect with the product leaders and ask your really specific questions. So, we will have breakouts on flow and automation, reporting and analytics, Agentforce, and low code and pro code development. And actually, a couple of these are going to be live on Salesforce Plus. So, it's a really great chance to connect with product leaders and just be able to attend that extended Q&A. Mike: Yeah. There's definitely a theme developing here across all three things that we talked about, which is access to the product leaders. Eliza: Definitely. I think that's a really key differentiator at TDX this year. I would say that there are just really more opportunities than ever for you to actually connect with the people building the features that you're working with every day. Mike: Now, one of the things is the virtual TDX Hackathon, which I know Josh on our team has been managing a lot. If you're not participating, that's something really cool to go see because I think we have... Correct me if I'm wrong, but I think we have an actual kind of showcase where they're going to get up and pitch their hacks. Am I right, Eliza? Eliza: Yes. And it's not just a showcase. It's actually even bigger than we've done before. So, it's going to be live in the keynote room. So, really getting everybody up on that big stage. And the three finalists will actually be pitching their builds live in front of a panel of judges. Mike: Wow. No pressure there. Eliza: Yeah, seriously. Mike: But I think that's really cool. Jennifer, if you were a customer and you're like, "Well, I didn't participate in the Hackathon," sometimes going to see what other people are envisioning might actually spark an idea for something that you're trying to solve back at your org, right? Jennifer: Oh, exactly. At Dreamforce, we had the Hackathon and I attended that. And it's always interesting to see how people solve for things and you're like, "Huh, I didn't even think about that piece, and maybe I should consider that and go back and use that in your implementation." So, yeah, it does definitely inspire and get those creative juices going. Mike: Right. On top of just watching people try to pitch, I think that's really fun too because I always love to learn watching other people speak because I would envision, well, here's what it could be like if I had to pitch my app to a board of directors or a president of a company as we're really going enterprise-wide as an admin. The last thing that I see is Agentforce Vibes, which I've had people on vibe coding on the podcast, and I've seen some dry runs. I can't tell you the session because you need to go to them, but there are some presenters in the admin track that have some really cool use cases for vibe coding. And so, I think we have the Agentforce Vibes Zone. Eliza, do you know anything about that? Eliza: A little bit. Yeah. So, it's an immersive hands-on space. And I think what's really cool about the Agentforce Vibes Zone and what we've done a really good job with is just making sure that it's not just for developers. I think no matter your role, you can definitely benefit. And so, it's really just designed for anyone that's ready to expand on how they build using natural language. So, there will be a series of Agentforce Vibes demos, and then, I think they're doing a guided hands-on project so you can apply what you've learned because as you know, if you've come to a Salesforce event, it's all about getting hands-on. But there's a lot of great vibes content across the event from sessions to this specific zone. So, I think if you're an admin and you're just trying to dig in or get started, it's definitely a great place to do that. Mike: Yeah, I'm excited for it. Let's talk about conferences in general, because it's been a while since we've done Dreamforce, and we've had winter and a time change. Jennifer, if you were a customer today listening to this podcast and thinking, oh, I got to pack for TDX, what are some things that maybe you bring now that you wouldn't have brought five years ago? Jennifer: Unless you really need to use it, I leave my laptop in the hotel just because lugging that around all day isn't fun. Mike: They still haven't gotten any lighter, have they? Jennifer: No. Mike: They're still 10 pounds. Jennifer: No, they feel like a brick, ton of bricks. I would definitely bring your most comfortable walking shoes. Don't bring the pair that you just bought. No, that's not a thing. Bring your most comfortable shoes. Definitely bring a water bottle. Walking around, you want to stay hydrated. Bring some snacks in your backpack because sometimes you think you have time to run and grab something and you're running between sessions. Have that little candy bar or whatever it is, a pick-me-up. Yeah, that's helpful. And then, I also bring a bag just in case you pick up any swag along the way because they'll- Mike: Just in case, as if it's an option. Come on. Jennifer: But bring a thing where you could take notes. So, whether that's electronic, or a notebook, or old school, bring something that you can jot notes or take screenshots of things. There's also will be available in the app, PDFs afterwards, but I always like taking a screenshot, and then, writing notes afterwards. Mike: Yeah, I definitely... So, there will be AI-generated session summaries on all of the sessions in the app, but being there and taking notes is usually... The notes that I take always make more sense to me, which is exactly how it should be, I suppose. Eliza, is there any special prep you do to get ready for TDX? Eliza: Well, I would say definitely pack your headphones because we have some sessions where you'll want your headphones to catch that session audio. Mike: Oh, that's a good one. Eliza: Yeah, that's one that people forget. You don't think about it, but got to have those headphones. I think for me, with these conferences, it's so good to have a game plan. I think when you roll up first thing in the morning, you just have this whole day in front of you, and it can be hard to think about what you'd like to do. So, I'd highly recommend using Agenda Builder to really plan out your perfect schedule, but I think you should also leave room for spontaneity because you're going to be connecting with people. You might see a cool activation that catches your eye. Sometimes there are robots wandering around that you want to stop and grab a picture with. Jennifer: Or characters. Eliza: Or characters. Yep, absolutely. So, don't jam pack it, but it helps to come in with a little bit of a game plan. Mike: Yeah. Dreamforce had puppies for a little while too. I know. Eliza: I missed those. Mike: We probably won't [inaudible 00:16:32] puppies this year. I know. One thing that I think is different that I'd like both your perspectives on. So, travel budgets are tightening. Companies are sending you to... Sometimes you have to pick and choose, do I go to TDX or Dreamforce? And sometimes it's, well, we can only afford to send you. We can't afford to send the other three admins. And so, you're there flying solo. What would your advice be, Jennifer, on making friends at TDX and introducing yourself? Jennifer: Oh, Mike, you asked the introvert of the team. Mike: I know. All of us admin evangelists are really introverts. We just play extroverts [inaudible 00:17:21]. Jennifer: I know, right? I tend to look for a familiar face, but also for someone who might be by themselves, because if I'm by myself and they're by themselves, why not connect? And then, you could be a team, and then, figure out, hey, what sessions are you going to? Because maybe they might be going to sessions that you didn't even think of. But also if you're attending a conference and your teammates don't go, I would take copious notes so that then, when you go back to your company, you can then knowledge share, right? You weren't there, but here's the things that I learned and here's the PDF to that presentation, and give them a little walkthrough of what they missed, but also we're going to be recording a bunch of sessions this year, right? Mike: Oh man, we're recording so many sessions this year. It's insane. You have no idea. Eliza, how about you? How do you make friends at events or how would you advise admins to connect with each other? Eliza: Oh, I consider myself a little bit of an introvert as well. So, I think if I'm in a specific area or session that is really, really interesting...
/episode/index/show/buttonclickadmin2/id/40773910
info_outline
What Is My Trust Center and How Does It Help Admins?
04/02/2026
What Is My Trust Center and How Does It Help Admins?
Today on the Salesforce Admins Podcast, we talk to John Maxey, Senior Product Manager at Salesforce. Join us as we chat about how My Trust Center can help admins communicate incidents, plan releases, and operate with transparency. You should for the full episode, but here are a few takeaways from our conversation with John Maxey. From static status pages to personalized trust When I was early in my admin career in 2008, keeping track of status updates was as simple as loading up Salesforce Trust. These days, with so many different products and services, it’s gotten a lot more complicated. That’s why I was so excited to sit down with John Maxey. He’s working on My Trust Center, a personalized, authenticated experience that only shows you information that’s relevant to your org. Reducing noise and creating clarity for admins With the new My Trust Center, you’ll be able to get more specific information about upcoming maintenance and how it affects your org. Everything is tailored to what products and services you’re actually using, instead of having to sift through unrelated incidents and interpret whether or not they apply to you. As John explains, Salesforce can be much more granular about any specific maintenance or updates and how they will affect you. And that makes it easier to make decisions like when to promote new features, when you need to do testing, or when there might be downtime. You can coordinate better with your team and avoid surprises. Greater transparency through targeted communication At its core, My Trust Center is about improving transparency both internally and externally. For admins, it will provide more visibility into what’s going on when something doesn’t work and when a particular service will be back online. If there’s an incident, you won’t need to ask your CSM or contact customer support to figure out what happened. RCAs will be attached to each incident, so the entire process is self-serve. And you can configure notifications via SMS, Slack, or email to keep your entire team up to speed. Make sure to listen to the full episode for more from John about what’s coming with My Trust Center. And make sure you’re subscribed to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce Admins Blog Post: Admin Trailblazers Group Social Love our podcasts? Subscribe today or ! Full show transcript Mike Gerholdt: This week on the Salesforce Admins Podcast, I am joined by John Maxey to unpack the evolution of the Trust Center into something far more than a status page. That's right. It's becoming a personalized command center for how you run your org. I mean, let's be honest, it's not just about uptime anymore. It's about how you design systems that balance data, automation, and AI while keeping your stakeholders informed and confident. We're going to dig into what My Trust Center means for how you communicate incidences, plan releases, and operate with clarity across increasingly complex environments. So if you're the person everyone turns to when something breaks or when they just think something's broke, this episode's for you. Let's get John on the podcast. So John, welcome to the podcast. John Maxey: Thanks, Mike. Thanks for having me. Mike Gerholdt: Well, this is exciting. So I like it when I can have a podcast and a blog post on the same subject in the same day. And all of this stuff around trust and security, I feel like admins are right there, we're the Sentinels every day. But before we talk about some of the cool stuff that you're working on, let's learn a little bit about John Maxey. So John, how did you get to Salesforce and what is the cool thing that you oversee? John Maxey: Oh, well, thank you for that. So I came to Salesforce in late 2007. I joined as a customer support rep in CSG in our customer success group. I had just come from the FinTech, and if folks remembered, FinTech was kind of not doing well at that time in 2007. So it was an opportunity to switch industries and come to Salesforce. And that was pretty exciting being the ... At the time, our moniker was no software, sort of breaking the mold of traditional software and sort of moving to the cloud. Mike Gerholdt: Yeah. We were talking before I hit record. I think if you do the math, our Salesforce experience is definitely in high school at some point. John Maxey: Seems that way. Mike Gerholdt: They graduated from high school, which is kind of scary at some point. John Maxey: I wish. Mike Gerholdt: I know. Well, it's about to head off to college and make questionable decisions and maybe be a paleontology as a major. Who knows? But we're going to talk about ... So the blog post that went up today on admin.salesforce.com is the same topic as this podcast, which is My Trust Center, which I can't tell you how excited I am for this. Because I remember back to my days early admin, like '08, just going to having to pay. It never dawned on me until one of my users couldn't log into Salesforce a long, long, long, long time ago. And I was like, oh, I wonder why that is. But I never thought about like, oh, cloud services can go offline. I need to pay attention to this. And at the time, trust.salesforce.com was literally just like a handful of blinky lights. That's all it was because there was a handful of pods that you paid attention to and you just checked to see if yours was up and that was it. And then over the years, I mean, there's so much more to pay attention to. Everything that goes into trust.salesforce.com. So can you kind of take us on that journey of like Mike from 2008, just having to remember if NA1 is up to everything that the new My Trust Center is going to give us? John Maxey: Sure. Yeah. I remember using the Trust site when I first joined Salesforce as well way back in 2007, and it has changed a lot. So if you think about it, in 2007, Salesforce was really sort of selling one product, CRM, Salesforce automation. We had, like you say, we had just a handful of instances where we had customer installations. So even the field of boxes that we had there was small enough for folks to be able to find out what was going on with their stuff. And even in the URL of their application, it would say which instance you were on. So fast-forward 18 years and we've acquired companies, we've developed new technologies and new products. Some of them are on the platform and others are what we call off core or they're not directly in our application. They're separate different types of architecture. And then we've also moved into the public cloud. We've also been selling to enterprise customers and we think about an enterprise customer and their implementations are far more complicated, multi-cloud. So they may have marketing and commerce, as well as sales and service. And then we have all our industries. So in talking to customers when I became the product manager of the trust site in 2018, just sort of realized that their experience, especially for our larger customers was harder to deal with. When you look at, I mean, I think we have 950 some odd instances in our sales service industry, we have over a thousand DBs in marketing cloud engagement. It's just kind of daunting to try and figure out where you are. And so at the time we were figuring out tools to make it easier to find that. But really the answer was to get personal and to create an authenticated site. So then we know our customers. I believe that's like CRM 101, know your customer. And then when we know who they are, when you log in, we can present you with what you have with the things you already purchased, the products and services. So in this first release, we're starting that path, that journey to true personalization. And so there's still some gaps here and there, but one of them is the support for all our products. So we have requests out to all of our different products like MuleSoft and Tableau to do a little bit of work to get them ready for the My Trust Center. But yeah, so that's, I mean, it's just grown and this My Trust Center really just has come out of customer feedback and wanting a better experience. And the other thing about the personal, I know a lot of folks are going to probably jostle a little bit at the move away from a purely public site to the authenticated site. But there's an opportunity there to not only provide the personalization, but also be able to expand the types of services we can provide, product communications, things like that, that we may do via email today, but we can provide an experience to consolidate all that into one single spot, seeing as we will know who those customers are, being able to target the communications. As if we did it today and it was public, it would just be too much. It would be information overload. And so being able to segment that because we know who you are is really a powerful driver for that. Mike Gerholdt: Well, if you think about it, so I remember going to a user group, this'll totally date me. But they were like, "Well, you can subscribe to the RSS feed for your server status." And that was it for me. I was like, "Dude, I'm set. This is great. And I'm going to get the emails. It's perfect." And then you kind of don't know what you're missing until you don't know what you're missing. And there was rarely a case, but this also speaks to why My Trust Center makes so much more sense. If I were to go there, "Oh, my instance is down, but these three instances aren't." Well, it's not like I could do anything. It's not like if my lights are out in the neighborhood, but my neighbor's lights are on, I can grab my dog and I can put my slippers on and walk across the street and be like, "Hey, can we sit at your house until our lights come on?" It never dawned on me when I would go to trust.salesforce.com and see, "Oh, well, everyone else is online, just this one pod is off." Okay, cool. Next. Right? And that matters now because now if you think about it from what admins, developers, architects, they have so much to pay attention to. And all of a sudden, especially in a highly regulated industry, you talked about being in Finserv, why is this down? And you can feel that user looking over your shoulder and then you having to sift through a trust site, multi-layers, multi-things, and kind of give them an answer. This is, why can't you just present me with the dashboard that makes sense? I mean, the dashboard in my car only tells me about my car. It doesn't tell me about every other car on the road. So it kind of feels like a natural evolution. John Maxey: Yeah. And that was definitely part of it as well. The information overload as part of a growth and when you have a company that sells many more products than you started with. And then also we find that sometimes somebody will see an issue and then they go into the trust site and they see another issue going on and maybe they attach themselves to that. And so that can kind of confuse things when they reach out and report that. If there is an issue, we'll have ways to report that. But you shouldn't just assume that because somebody else is having an issue that it's the same issue that you're having. So yeah, definitely making sure that we're giving you the information. And then also being able to expand on it more. Like I mentioned, the public nature of that, we've got press watching it, we've got competitors and things like that. And sometimes we just don't want ... We don't have to be guarded with what we say because of how that could be used. When we know we're talking to customers, we can be a lot more open, even drafty as they might say. We know that something's going on. Maybe we don't know exactly what it is. And that's not really necessarily appropriate when you're speaking to a public audience. But when you're talking to customers and they just want to know that you know that there's a problem and that Salesforce is working on it. Like your analogy, when the power goes out, typically the customers don't need to know when the application is having an issue. They know that the application's that. They want to know that Salesforce knows that there's an issue and what are we doing to fix it? Mike Gerholdt: Yeah. I mean, ironically, it's just like the power company. John Maxey: It's just like the power company. Mike Gerholdt: So I'm looking at the screenshot and the other thing I noticed that I think is really neat that before back in the days of the blinky lights, it was like, it's on or it's off. Cool. And then we would pay attention to releases and you always got like a release date window. This shows a tab upcoming maintenance for the next 90 days. Tell me more about some of the thought that went into that. John Maxey: Yeah. So we look at the different personas that we want to support on the trust side, on My Trust, and we think about the jobs that they need to do. And so you have an admin. Admin is obviously concerned about any incidents going on to make sure they can communicate with their user base. But they're also working on developing new functionality for their user base. And so they want to know what's coming up, what would impact them. And in the past, it would have been as much of how much downtime would there be, but that's almost a thing of the past. But really it's more about when will my features? Maybe I've been working on those in Sandbox and I'm ready to promote them. When do I know to promote them? If we're talking about an operations persona, they need to know when there need to be a little more diligent because there's a release going on. Some of our customers do testing after we do a release to make sure that there's no conflicts or anything. Obviously as people customize their orgs, we do our best at Salesforce to make sure that there's no conflicts in the code that we release, but there's always, folks are very inventive in the way they do their solutions. And so there could be a conflict. A lot of customers like to test that, they want to know when. Especially if you have a large application, a lot of people in to test, you want to know the best use of their time because it's usually on a weekend. Mike Gerholdt: I know. Yeah. And I mean, even scrolling down, we could talk more about this too. I'm looking at a maintenance window of four hours. I remember before we would get a release as an admin, I'd always tell my users, every now and then that scheduled maintenance window would come up when you'd log in and be like, "Okay, well, just don't log in this weekend." And it's kind of like it's fun to think of the days when companies would only work Monday through Friday. Please salespeople just ... I know you're not working on Saturday, right? Wink, wink. John Maxey: Right. Well, a lot of the ... The major releases now, especially in Hyperforce are zero downtime. It's just really, it's more of a notification, "Hey, this is when your features would be ready," as opposed to it actually being downtime. The first party, it does have a little bit more. But I think we've got most customers off of first party now. And then we have some others that we're still working with to get their messaging better and because now we can target. So those timeframes, those windows will start to shrink because we'll be able to target folks at a better granularity and make sure they know exactly when they're going as opposed to a more generic sort of announcement, "Hey, we're doing a bunch of work. You may or may not be impacted in this time." We'll get it down to, we know when you're going to be deployed and exactly when the impact may happen. Mike Gerholdt: Yeah. And this is also going to help our inboxes because they're not full enough. John Maxey: I hope so. Mike Gerholdt: I know. Man, I'm telling you. The one line, say goodbye to instance wide email blast. Well, those were just fun. Okay, I think it ... Does it mean us or not? But still, one of the features is you can still subscribe to notifications, right? John Maxey: Yep, exactly. Yeah. Today we support email. We're looking to expand that in the near future to support SMS like we do in the legacy status site. And looking to add some more as well. I mean, obviously Slack is top on our list to investigate how to do that in a meaningful way so folks can subscribe to their notifications. Mike Gerholdt: Oh, like a Slack notification? John Maxey: Yeah, exactly. Like a Slack channel you subscribe to and get your notifications right in Slack. I mean, another one of our goals in our group in general is to try and push our information into the workplaces where folks are working. So if you spend your time in Slack, why not get that stuff in Slack? And then it's more meaningful and you don't have to switch to a different application. I mean, I love developing the trust site and the portal. But the main goal is to make sure folks are informed and get what they need. Mike Gerholdt: Well, so that was exactly ... I mean, to kind of ask you a best practices question, it could be, and in my case, I was the only admin, so it was easy for me to subscribe to everything. But there are a lot of companies that have teams, right? There's admins, developers. So putting it to Slack, obviously then it's anybody can join that Slack channel and get the status update. And if you need to know that, then join that channel, that makes sense. In the interim, how do you address that as a best practice for organizations with maybe multiple admins or multiple developers? John Maxey: Sure. That is a challenge. I mean, the quick and easy answer is we use Trailblazer ID, which is the same authentication method for Trailhead or for help, for AppExchange, all of our external facing portals. So there's a good, better experience there and one authentication through all those portals. But then because Trailblazer ID just takes an email address, you could use your group email to sign up for a Trailblazer ID. And then that ID, that email address, would have to be attached to a license in the orgs you want to see. So that's the way to do that. Today, we understand it's a challenge. Not everybody has a bunch of extra licenses to use for this type of purpose for folks who are monitoring. So we are working on a solution for a license that would have very limited or maybe no access to the org at all. But you could attach it to say a group email and then be able to get all the monitoring information you needed from that. Mike Gerholdt: I like that. John Maxey: Yeah. Mike Gerholdt: That's awesome. It says one of the plans is to have major release notes and root cause analysis for eligible incidents. How is that beneficial? Like what were customers asking for that we weren't providing root cause analysis on? John Maxey: As much of anything, it's a delivery mechanism. So if a customer is impacted by an issue, they may have to engage with their CSM or go to customer support and say, "I need an RCA." And so instead of that, we have certain criteria on incidents when an RCA is going to automatically be provided. So our first iteration would be that when that RCA is provided, we would attach it directly to the incident and you would get a notification that it's been attached. And so you could then go reference it. No need to ask anybody. In the future, we're thinking about ways maybe we could provide some visualization to say that it's in process and maybe how far in the process. I mean, the more we can self-serve, the better. And then we'll look at other ways of being able to surface that type of information and similar type of communications directly in the portal and through notifications and all of that. Mike Gerholdt: Before we press record, you said some customers already have their hands on it. What was some of the feedback they gave you? John Maxey: Oh, we were in Pilot for about a year with a handful of customers, and then we were in beta for ... Our open...
/episode/index/show/buttonclickadmin2/id/40698675
info_outline
How Salesforce Admins Are Evolving to Run the Agentic Enterprise
03/26/2026
How Salesforce Admins Are Evolving to Run the Agentic Enterprise
Today on the Salesforce Admins Podcast, we talk to Rebecca Saar, Senior Director of Admin Relations at Salesforce. Join us as we chat about governance, productivity, and why Salesforce Admins are more essential than ever in the agentic enterprise. You should for the full episode, but here are a few takeaways from our conversation with Rebecca Saar. Admins are becoming the guardians of trust As AI becomes an ever bigger part of our day-to-day lives, I wanted to bring Rebecca Saar on the pod to talk about what changes and what stays the same. One thing she emphasizes is that admins are here to look at the big picture for their organizations. While it might be easier to build the solution, you still need to figure out what to build in the first place. “It’s a shift in mindset and understanding where the work is happening,” Rebecca says, “because we now can leverage these super powerful tools.” Skills like gathering requirements, talking with stakeholders, and bridging the gap between departments are only going to be amplified by AI—not replaced. Admins act as translators across systems and teams Admins are a crucial intermediary for their organization, connecting the dots between multiple departments to understand what’s really going on with a business process. In other words, admins are the ones who understand who to talk to in order to get something done. In my experience as an admin, each business unit only knows their own siloed data. They know where it comes from and who they ship it off to, but they don’t know if there are better options, or where there might be a snag in the process. That’s where admins need to step in and find a solution that individual business units might not be able to figure out on their own. Strong foundations still matter in an AI world Despite new tools, core admin skills remain essential. In a way, we’ve moved from problem solver to sense maker. You might not have to spend as much time figuring out who has what permissions, but you’ll still be called on to have a conversation about how it should be handled. Just like with everything else, AI can save time on the busywork so you can focus on what matters. At the end of the day, it’s even more important to lean into the foundational admin skills to find solutions that work for everyone in your organization. Make sure to listen to the full episode for more from Rebecca Saar about what’s coming at TDX, and don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Blog Post: Salesforce Admins Blog Post: Salesforce Admins Blog Post: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins podcast, I sit down with Rebecca Saar, the Senior Director of Admin Relations for Salesforce, to talk about how the admin role is changing as AI and agents become really part of our everyday life. Now, we're going to dig into why Salesforce admins are becoming the guardians of trust, how strong fundamentals still matter, and where human judgment fits in when automation gets smarter. I like to think this is a very thoughtful conversation about governance, productivity, and why Salesforce admins are more essential than ever. So with that, let's get Rebecca on the podcast. So Rebecca, welcome to the podcast. Rebecca: Hi, Mike. Thanks for having me. Mike: It's been a while. Last time you were on, we did an entire episode in German because international podcasts are fun. Rebecca: Yeah. What a ride. How many years ago was that? Mike: Just a few. It's like one or two in Salesforce admin years, which translate differently to the rest of the world. Rebecca: Right. And then I also say we've had a pandemic in between, so that has made time fly. Mike: Yep. Rebecca: And I think that was a pre-pandemic thing. Mike: It was. 100% pre-pandemic. Yep, absolutely. And then since then, you've been on stage quite a few times for the admin keynote. And doing a whole bunch of stuff with the admin relations team. You and I have both been on stage too. Rebecca: I know. Yes, we have been championing admins since 2014? Mike: Since forever. Rebecca: Yes. Dedicating over a decade and celebrating that big decade 10-year anniversary was a couple of years ago. And that was a big highlight for me, being able to share that with everyone on stage at Dreamforce. But yeah, since last time I was on the pod, I have now become the lead of admin relations and kind of leading this amazing team of evangelists and marketers on our mission to enable and empower and inspire all of our Salesforce admins out there. Mike: And we're doing that as a transition to talk about... Literally right before I pressed record on this podcast, I was listening to NPR talk through AI and how AI is affecting something. They quoted 60 to 80% of the people in the workplace now use AI for some sort of task. And it kind of flashed back to me because it was interesting listening to them talk. They were describing, so when you tell AI to do something, that's a prompt. And listening to people walk through the understanding that I think a lot of Salesforce admins went through two or three years ago, I remember having people on the podcast to talk about prompt engineering, and I'm listening to them now, I'm like, "How do you not know this already? What world are you living in? You should already know about prompts and prompt engineering." But leading admins now as we're not just working with AI, but AI's working with us, right? Rebecca: Yeah. AI is everywhere. I think just in the last few years, it's accelerated so, so much. Yeah, I've been thinking about this a lot because not only are our admins and our whole Trailblazer community thinking about the impact of AI. But it really is across at least tech industry, but corporate world where everyone is kind of thinking about how are we incorporating AI into our workplace. Mike: So I was thinking of the article that you published, How the Salesforce Administrator Role is Evolving in the Agentic Era. And the first kind of big bullet point or the first subhead under there is The New Mandate, The Guardian of Trust. And working on that article with you, I happened to think of the very first thing that I did when I was a Salesforce administrator. Well, the second thing I did. The first thing I did was I updated my profile. The second thing I did was I got certified and I was one of the first 500 Salesforce admins certified. And I remember posting that to... We had a local kind of intranet at the company that I was at. And I remember posting that just out of pride for myself that I had passed, but also really wanting my users and my executives to know like, "Hey, when I suggest something, it's because I'm knowledgeable about it, so you should trust me." And I think it's interesting that almost 15 years later, when AI comes around, this is the very first thing that we're talking about with admins as being the guardians of trust. Rebecca: Yeah. Trust, it's our number one value. And it's, I think, the most important thing our admins can do for their company and their users. You manage the data, you manage security, you manage automation. And these are the pieces of the system that need to come together in a trusted way. It makes the admin critical to the success. Mike: A lot of it is also helping our users and ensuring that when we implement Agentforce for a specific process like in sales or in service, that it's really something to aid them. I think of a lot of people, the discussion on NPR was, "Well, I don't know what to ask or I don't know what to do." And of course, as an admin, when you're demoing something for Agentforce, you want to show, here's the words you type in. And you'd like to think in a perfect world, all of your users are going to use that exact perfect prompt when in reality they're not. People forget prompts, people forget what they're trying to do, and then they can easily get frustrated. I think part of that guardian of trust that we're doing is, let's put it in places where it can benefit our users the most in terms of productivity that might be easy to just farm out so that really we're bringing the best out of that person. I think of now when I was deploying a contract object a long, long time ago, the people that worked in contracts, the value of them wasn't the fact that they knew how to open an envelope or read through a paper contract or photocopy it. And I say that because at the time we did all of that, it was all physical contracts. That's not what they were getting paid for. They're getting paid for their knowledge of being able to digest that information of the contract. And I think of kind of the same value for the admin. It's not the buttons they click that make it important. It's how do they implement the agent and then how do they use their judgment to say, "Let's put this agent here at this point in the process to make all of our users really use the skill that they're being hired for." Rebecca: Yeah, 100% percent. I really feel that these tools are coming to help us be more productive. We keep kind of hearing this narrative around with AI that we're replacing roles, we need less people, and does that mean there's room for my role? And I think that's really missing the big kind of upside of all of this is that there is so much work to be done. Especially when I talk to admins, they're wearing multiple hats. They've got cues and cases that they're trying to get through. And think about all the ways that AI can then help you as an admin be more productive and get more work done, which ultimately means you can serve more of your users and the company. Mike: Yeah. I think of it as... So I was trying to think of other industries that we've seen major shifts in, I think AI is really close to automation. And if you look at how auto manufacturing has evolved from the moving assembly line of the 1930s and '40s with Henry Ford up until a current day, I want to say, and I've watched a couple of YouTube videos because I think it's fascinating, but realistically a car when it's being built isn't touched by a human for the first something like 10 hours of it being built. And it only takes like 30 hours to build a car, which first of all is kind of insane when you consider how much there is that goes into it. But even if you look at stuff that would be highly profitable to be fully automated, we still haven't fully automated anything because I think even as humans, it's that level of control that we need over things to make sure. But the parts that like for the car being built, for example, we're automating is, okay, well, we can do this repetitiously. And it's a very low level skill as opposed to really hiring people. There's still people you think of 100 years later that build cars in a car manufacturing plant, robots never took over. It's just the part that they get hired to do is the very highly specialized part. And I think of that with AI where it's not going to replace Salesforce admins. It's the admins that know how to use it are going to be the hyper productive built out systems because they know when to have AI built for them and when they need to jump in. Rebecca: Yeah, exactly. I think the work shifts from the building piece more to the before planning and preparing, and all that work admins do talking with stakeholders, gathering the requirements, really getting precise on the solution. And then the work after the building where we're monitoring the agents and assessing how they are performing. It's just a shift in mindset and where the work is happening because we now can leverage these super powerful tools. Mike: Right. There's parts now, and we've even seen this. I've had Cheryl Feldman on the podcast to talk about how setup is getting better. And if you're watching any, I think this last... Well, when I recorded this is in March, I just watched Mo do the Agentforce Now tour virtual workshop and walking through agents. And there's still parts of it where the admin needs to understand what the part of the business is that I need to go to talk to, because a casual user probably doesn't know that. And then also, how is this stitching all of the data together? Because as I found when I was an admin for eight years, everybody in an organization only knows their own siloed data. And they only know where it comes from and who they ship it off to, but rarely do they know if either of those are exactly the forms of data that they need. And so for example, like when I was doing a sales implementation, the information that would come in from the opportunity, the finance team was like, "This is of no use to us. Somebody else needs this. Here's what we need." And that's the part where actually sitting down and making those judgment calls as a Salesforce admin is, "Okay, here's how the agent can help do that." The two operating divisions might never come to that conclusion on their own. Rebecca: Yeah. It's really the admin running the system. There are different departments, there's different tools we're using, there's data in different places. So how does it all come together and how are we looking at the day-to-day operations of these systems and maintaining them and scaling them? Mike: Now, I think what's interesting, Rebecca, so that article was super fun to help write and get out. This was the beginning of a journey, and I'll link to it in the show notes. You also spoke with a couple of next gen admins, we'll call them that, with Tony and Ishrat. I think so looking through those blog posts as well, what were some of the things that kind of stood out to you? Rebecca: Yeah. First of all, I'm so grateful to all of the admins I've been talking to in the last couple of months. Through those conversations, it really helped define that blog, Mike, that we wrote together, because there's so many insights that kind of came to light and a lot of grounding and agreeing on areas where it's important to have focus. I think one thing that really came to light in Ishrat's was talking around governance. And really kind of thinking about the admin as going from problem solver to sense maker, which I thought was an interesting phrase. But as the admin, you are kind of the translator of your org. You understand or need to understand what all these pieces are and how they work and why an agent is doing what it's doing, because you kind of helped bring it to life in the beginning. So that was kind of an outstanding insight there. And then from Tony, I had some great conversations around, I think a highlight for me was just thinking about the importance of the foundations and knowing the kind of foundational skills of an admin. And not losing sight of that as we leverage and use more and more AI. Still knowing what a data model is, knowing how to set up a flow, how these core kind of building blocks work well so that you can be the sense maker and understand eventually what an AI is doing and automating for you. Mike: Yeah. I think looking through both those, I love the way they ended. Tony kind of sums up, "I'm going to give you what you need, not just what you want." And essentially Ishrat says the same thing of the person who connects business intent to data quality. And I think that's always the thing. There's always the... And I used to explain this a lot when I would do training, you run the report and you think you know what it's going to say. And then the data comes back with something different. And nine times out of 10, the user would be like, "Well, the report's wrong." Not that you put the data in wrong, the data is telling you something different than what you thought. And I think with both of them, that's the role that the Salesforce admin plays so critical is, "Well, here's the intent of what we're trying to capture and here's the agent capturing it and spelling it out for us. Now, here's actually what the outcome is and maybe the business process that we need to change as a result of it, as opposed to what we were focused on. Rebecca: Yeah, love a good reporting situation. How can we change this report? Because it's not the outcome I was looking for. Mike: Right, exactly. I need more pie charts. That's always the end result is the pie charts. But this is really cool. I think right now all of the roles are really sitting back and trying to understand how they fully utilize this new tool, because I think of it like how Steve Jobs explained the bicycle of humans can move, they can put one foot in front of the other, but the bicycle makes us infinitely so much more productive. And I look at how AI is doing the same thing for Salesforce admins in terms of just allowing us to... Some of the work that Cheryl's doing was set up, asking query who has the right permissions, as opposed to spending that time trying to run reports or comparative reports, having it do that work for us so that we can sit down and have those really targeted conversations about who should have the right profiles and permission sets, as opposed to spending that time just looking through thousands of reports. Rebecca: Exactly. It's an exciting time. I really think obviously we sit in the bubble here and we see all the innovation firsthand. But it does feel like a very exciting time in our industry and for Salesforce and for the Salesforce ecosystem. There's so much opportunity for those that are excited and willing to kind of embrace this and bring this to their companies. Mike, I was curious how you're feeling. We see this incredible opportunity and the value the admin brings having talked to admins out in the community, is that resonating? Are people- Mike: I remember sitting down at Florida dreaming not too long ago, and listening to all of the Salesforce admins that were not only paying attention to what Agentforce could do, but what other AIs can do for them. I think you mentioned we live in a bubble and I think in terms of my friends, because I work in tech, I definitely know more about AI than most of them. But I would consider myself medium level intelligence of AI in the tech world. It's really, this is another tool, how do I be productive with it? And they're not looking at it as, "Oh, this is something scary." They're looking at as, how do I embrace this? And I think it's the same way that we look at how computers came about. Once upon a time, as I've seen on watching Mad Men on AMC, people typed. There was huge steno pools. If you needed a copy made, somebody literally had to type copies of that. And then the copy machine happened. Well, the copy machine didn't put people out of work. It just made people more efficient. And so companies were able to move those individuals around into places that could make them more efficient. And admins now have a tool that just kind of, it's like literally going from walking to riding a bicycle. And of course, there's lots of different bicycles and everybody has their own preference. Rebecca: Exactly. I feel like the admins who lean into this won't just be staying relevant. They're going to be the most important people at their company. Mike: Well, and as I was having the conversation with Daniel Peters talking about small business, if you think of, "Well, I'm just a Salesforce admin at a small business, what can I do?" He actually brings a point up of, if you built an agent that would literally just help the person answer questions while they're filling out a form or while they're creating a record in Salesforce, you would have so much better data. And his example was when they would take orders for cheese over the phone, there was insane amount of information that you had to know. This certain cheese could only be sold in these sizes and this certain cheese could only be sold during these months. And for the person, the on-ramp, you think...
/episode/index/show/buttonclickadmin2/id/40619000
info_outline
How Can Agentforce Help Manage a Salesforce Backlog?
03/19/2026
How Can Agentforce Help Manage a Salesforce Backlog?
Today on the Salesforce Admins Podcast, we talk to Neil Foglio, Solution Architect at Arkus. Join us as we chat about how to use a backlog to prioritize requests in your Salesforce org, and how Agentforce can help improve transparency and trust. You should for the full episode, but here are a few takeaways from our conversation with Neil Foglio. Why Salesforce Admins need a backlog We talk a lot about forward-looking statements on this podcast, but what about that thing trailing along behind you? That’s right, it’s your technical debt. If you’re facing a mountain of requests, this episode is for you. That’s why I sat down with Neil Foglio. He explains how your backlog can be a tool to streamline your Salesforce org and help it evolve with your business. “It's not a to-do list and it's not a task list,” Neil says, “it's a commitment that something will be prioritized amongst everything else that you have on your plate.” It’s a central place to capture all requests, improvements, and ideas, and then evaluate them so you can figure out what to work on and when. Writing better backlog items For Neil, the key to making a backlog that works for you is to write better action items. Make sure that every item starts with an action verb. Instead of “create a new fundraising homepage,” be more specific: “design a new Lightning page so gift officers can see donation history.” The other key step is to evaluate and prioritize your action items. There are several different systems you can use, but you’re generally looking at value, effort, and risk. How much impact will making this change have? What resources will you need to get it done? And finally, what data do you have to support what you believe? For many organizations, there will also be custom dimensions that have their own score. For example, an educational institution may evaluate everything in terms of how it improves student learning. What’s important is to establish a clear process and get buy-in from leadership over how things are prioritized. How Agentforce can help with backlogs Managing your Salesforce backlog is even easier with Agentforce. As Neil explains, getting a clear and specific user story is crucial to creating useful action items. You can set up an agent to take requests, and go through the granular details with the user of every action they’re taking in a business process so you can get to the why behind their request. Agentforce can also help you improve transparency by explaining the why behind your prioritization decisions. You can have a helper explain what you’re working on and when you expect to deliver it. Be sure to listen to the full episode for more from Neil and hear his upright bass playing. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Salesforce Admins Blog Post: Admin Trailblazers Group Social Full show transcript Mike: Welcome to the Salesforce Admins podcast. And hey, you know what? That intro music you hear today, that's actually our guest, Neil Foglio, playing the upright bass. Neil is a solution architect at Arkus. And in this episode, we're talking about something every admin eventually runs into the backlog. Neil shares how thinking about your Salesforce org, more like a product than a project, can change how you manage requests, prioritize improvements, and make sure the right work gets done at the right time. We also dig into practical ways admins can evaluate ideas using simple frameworks like RICE so decisions feel fair and transparent. If you've ever wondered how to handle the constant flow of, "Hey, could we add this request?" This conversation's going to help. So with that, let's get Neil on the podcast. So Neil, welcome to the podcast. Neil Foglio: Hey, Mike, thanks for having me. Mike: I'm excited to talk about this because a lot of things on the podcast we always talk about are future-facing. And I feel like this topic is kind of future-facing, but it's also managing, well, I don't know how best to put it, the burden of debt that we somehow get ourselves into. And in that, it's backlogs, which doesn't sound interesting, but I promise you, you're going to be excited by the end of this podcast. So Neil, let's start off with introducing you to the world and kind of let people know what you do and how you got into the Salesforce ecosystem. Neil Foglio: Yeah, of course. Hello, everyone listening. My name is Neil and I'm a solution architect at Arcus, Inc, where I work with both nonprofits and enterprises implementing Salesforce. Most of my work lately has been helping organizations treat their CRM more like a product than a project, and to rethink how they steer their Salesforce org with their evolving businesses, where governance kind of comes into the picture, and that's where this topic of the backlog fits in. Mike: Yeah. Wow. That's interesting because I'm looking at the core responsibilities we have for admins, and project and product management are both core responsibilities. Neil Foglio: Yeah. I feel like a lot of admins now are really acting more like product managers, steering their Salesforce org, figuring out what elements have the most value and where you can produce the most ROI. And this is where you want to have systems that prioritize those things, and that's exactly what the backlog is. Mike: Okay. So I will confess this, that when I started as an admin a long time ago before the iPhone, we'll just say that as a date, I had no concept of some of the technical things that my developer friends knew. I didn't know about change management. I didn't know about backlogs or managing technical debt. For newer admins that are just getting started in the ecosystem and at their organizations, how do you define what a backlog is and how do you define what gets on a backlog? Neil Foglio: Yeah, absolutely. And I was in that same boat. So I started my tech career coming from healthcare. And so the first time I heard the word backlog, I thought it was a terrible thing, that something wasn't happening, work wasn't getting done, patients weren't being cared for. But when we think back to starting in a Salesforce org, you might have been in a position where you inherit an org or your company gets some new leadership that wants to make a bunch of improvements to the org and you find yourself questioning, "Well, where do I start and how do I prioritize all of these questions and features coming at me?" Without a system to rate them all, the things that typically will get done are those asked by the person with the most seniority in the room or the person that's emailing you three times a week. Mike: Right. Sometimes it's the same person. Neil Foglio: Could be. And sometimes it is especially when there is new leadership or a disruptive factor happens in the business that we need to respond to. And what the backlog is it's a tool that creates a level playing field for all of these ideas and their stakeholders. So it's a central place to do two things. Number one, to capture all of these requests and improvements and ideas, and number two, to rate them against each other using the same evaluative criteria so you can figure out which ones have the most value at this moment when you have capacity to build something new. Mike: That makes sense. I mean, to be fair, when I first started, the requests would... I would manage them basically in the order that they were received. And I very quickly had to learn, "Oh wait, creating a list view is way less effort than creating a contract management application," for example. And just because the list view is the third one down doesn't mean I can't just bump it up and knock out a whole bunch of half hour jobs as opposed to, "Well, it's going to have to wait until I get this monster of a contract application created." Neil Foglio: I mean, those little things tend to add up too, right? So if you're spending a lot of your time in a reactive mode and you're doing these requests as they're coming in, you might not really have a lot to show for it at the end of the day. Mike: You can. Yeah, absolutely. I had to learn how to manage my backlog. And I will say this, your backlog isn't your email inbox, right? Neil Foglio: No. It's not a to do list and it's not a task list. It's more of a commitment that something will be prioritized amongst everything else that you have on your plate. Mike: Sure. How do you suggest admins manage their backlog? Neil Foglio: Yeah. So if you were to Google this online, you'd see there are a bunch of different prioritization techniques. So before we get to prioritizing the items on the backlog, we should talk about what makes a good backlog item because those vague items like creating a new fundraising homepage are not something that can be easily estimated, right? So the idea here is that if you have enough detail in your backlog items, you can figure out how much time it will take, how many users it impacts, how much confidence you have that it's actually a problem that can be solved, and how much effort in terms of time, hours, or people will take to get it done. So the first kind of step is making sure that all of your backlog items start with an action verb. So using that fundraising homepage as an example, we can reframe that to say something like, "Design a new lightning page so gift officers can see donation history." In that one sentence now, we have exactly what we want to make. Maybe there's some reporting or dashboards in there too, who it's for, and what the impact would be for the organization. Mike: I like that. I think writing titles out and using different words or verbs or tenses is so my jam because when I was in sales, I would always write all my to do lists in the third person. So it was like future Mike telling present Mike what to do. And I remember all of the salespeople would make fun of me. I was like, "Yeah, but when I go to pull up my to do list, it literally tells me what I need to do right now and I don't have to decode what I was writing my notes on." So I like that you start everything with a verb. Neil Foglio: Yeah. Bring it into that trusted system, right? Mike: I mean, the backlog's there, the second you get 10 requests and you fulfill six of them, well, now you have a backlog of four, and then tomorrow you get 10 more requests. How do you balance between the two in terms of managing new requests versus shoveling all of the new stuff into the backlog? Neil Foglio: This is where the prioritization comes through. And so if we were to Google backlog prioritization techniques, you'll see a whole bunch of different methods, but what they all kind of have in common is that they convert three things: value, effort, and risk into a numerical score. So teams can compare those ideas more objectively. In doing that, you're not trying to get to an absolute value. Precision isn't really the goal, but again, we're comparing the relative priority. So what you'll see is a really common prioritization technique has the acronym RICE, where the R stands for reach. Think number of users. Let's say you have a defect and it affects a hundred users or a thousand external stakeholders who get a specific document generated. I is impact, which is usually a scale where higher numbers have more impact. The C is for confidence, which accounts for how well you have data to back up what you're saying. And this is typically given as a percentage. And E is the effort, and that is measured in time or resources. But whatever the technique, the most important thing is that this process of entering things into the backlog and revisiting the backlog happens on a regular cadence. So maybe it's appropriate if you have a pretty heavy backlog that it happens twice a month or once a month if there's not too much on it. But this consistent way to evaluate requests lets the prioritization process feel fair and transparent. And you can't put a price on that kind of trust, right? Mike: Yeah. No, that's good. So from a technology standpoint, how have you managed your backlog? Do you use cases in service or built a custom object or managed it outside of Salesforce? Neil Foglio: Well, I think I have to say that everything should be in Salesforce, right? Mike: Yes. Neil Foglio: Even if it wasn't on your podcast [inaudible 00:11:45]. Mike: If it's not in Salesforce, it doesn't count. Neil Foglio: Right. Exactly. But if you're a new admin that's starting out trying to figure out what needs to happen in your Salesforce org, you can design a custom object that has some picklist values, new in progress. You can create the custom dimensions you want to rate your backlog items on. So by custom dimensions, we just talked about reach, impact, confidence, and effort, but some sectors will have more tailored solutions. So for example, healthcare usually has a score that has something to do with patients, like risk of patient harm or impact to care coordination. Finance orgs are probably interested in reducing the likelihood of fraud or improving regulatory compliance. And I'm working with an educational institution right now, you'll think you'll find this interesting, that's evaluating every Salesforce idea on how well that idea improves student learning. Mike: Ooh. Neil Foglio: And we think about putting that all in a Salesforce object, you can have all those ratings right there as your picklist values so you can easily calculate a formula to say, "Here is the score of this backlog item, and here's how that compares to everything else that's currently in flux." Mike: Wow, that's really cool. I like that. How much... We've kind of skipped over just the request part of building a backlog. But when I was an admin, I think I always struggled with the ticketing system that I created, not in managing it, but in how many questions and how many fields do I have users fill out for a certain request? Because there's always that like, "I need a lot." And I remember a long time ago, Gary Palmatier told me, "Always ask people to tell it to you in plain English, not in Salesforce speak." Right? Neil Foglio: Right. Mike: And by that he meant, "I need to manage my incoming leads so that I ensure everybody gets called back 24 hours after a trade show and I send them a brochure." Not, "I need to manage new lead records and have your users describe what's going on in Salesforce." Where do you find there's kind of a sweet spot in terms of getting that information from users or managing that so that you can adequately decide, "Here's how much effort this one's going to take versus this one, which I'm going to put in the backlog"? Neil Foglio: That level of detail is really important because that allows you to estimate the effort. We wouldn't ask our users to tell us how long it would take us to implement a feature. And I think you're absolutely right that giving it in plain English lets you figure out what that user story is, and then maybe also lets you connect with people who can give you more guidance as to what it would take to bring that element to fruition, right? So I think getting that user story and that journey down from the person that's submitting the backlog request is probably what's most important because you'd want to know how do you replicate this? If it's a defect, what do I have to do to find this defect and resolve it? If it's a journey, if it's creating a document or sending out tax letters for donors, you want to know what those steps are so you can properly estimate it. And sometimes that comes just from doing and having your users tell you exactly what it is they are doing and what they'd want to do. Mike: Right. No, it's good. So Agentforce. Neil Foglio: Agentforce. Mike: Agentforce can help us do a lot of things. What do you see as use cases? Because this is also a great area for admins to build agents to help them manage their implementation. What do you see as use cases for admins to build agents to help with a backlog or with feature requests? Neil Foglio: Maybe the first question is, is building the agent on your backlog? Mike: Ooh, look at you going Meta. So is it people? There's the question for you. Neil Foglio: So is it? Well, I think when we think about what the agent's capabilities are, if you're already managing your backlog in Salesforce and not an external ticketing system or an Excel sheet, you can have your agent answer questions as to what similar backlog items have happened in the past because that's probably a really good use case because what happens over time with these backlogs is that they become this form of organizational memory. When you complete a backlog, you're not going to delete it from your backlog, it would be archived and completed. And that's when you start to see patterns in the requests that people make. And that's where AI is really useful in telling us what those patterns are. Mike: Yeah, absolutely. And I was also thinking too, boy, you could... There's a lot of... I don't know. Sometimes I have too many ideas and my mouth is going like, "What if it asks you had a question agent before they put in their request like, 'I'd really like to do this blah, blah, blah report'." And instead of just submitting the ticket, the agent's like, "Oh, well, that report already exists." And then it shows you the report and then it's like, "Yay, this agent saved me 18 tickets this week or something." Neil Foglio: Yeah. Or, "We're already working on this report for you and we expect to have it delivered in this quarter." Mike: Yeah. I mean, the agent could immediately look at the backlog and say, "Well, actually a similar ticket has been submitted and its status is X. So do you want to submit another one?" I think that was always something that I ran into, was one person have a good idea and then somebody in another department would have the same idea, but they think they're both the first person to have that idea. And so then you'd end up getting two or three tickets that it's like, "Well, these are all the same ticket," you know? Neil Foglio: That's where the backlog also is this collaborative tool, which if done really well, fosters this incredible sense of trust across the organization because you want to share those things with people who are interested in them. Mike: Yep. So let's talk about sharing. I created a dashboard that everybody in my org had access to. So they could see the number of tickets, they could see the status of tickets. And I felt like for me, the transparency kept everything honest. Somebody never said, "Well, it's just submitting it to a black hole." It's like, "No. When you submit it, look, the dashboard changed. Here's your ticket." What level of transparency do you suggest for admins in terms of sharing their backlog with their users? Because I think it can be tough sometimes when you have to make that decision of, "I need to pause on this because I can create these four list views and knock four tickets off my backlog as opposed to spending the next four and a half hours and creating this app or something because it can feel sometimes like favoritism." Neil Foglio: Oh, absolutely. And I think that's where backlog management isn't really a single person's job. I think the most effective teams usually have a committee, like a steering committee, that will review the backlog on a regular cadence together, which pulls in the executive sponsors of certain departments to make sure that the prioritization is happening in a fair and transparent way. And that level of commitment from leadership is pretty important for an initiative like this in large and small organizations because the backlog is your holding place, but it's also your prioritization space. And so a lot of people tend to get involved. Everyone wants to push their items up on the backlog -- Mike: Oh, yeah. Neil Foglio: ... unless their stuff...
/episode/index/show/buttonclickadmin2/id/40526825
info_outline
The Future of Salesforce Setup Is Agent-Driven
03/12/2026
The Future of Salesforce Setup Is Agent-Driven
Today on the Salesforce Admins Podcast, we talk to Cheryl Feldman, Senior Director of Product Management at Salesforce. Join us as we chat about how Agentforce is reshaping the Setup experience. You should for the full episode, but here are a few takeaways from our conversation with Cheryl Feldman. From 1,300 pages to one conversation Let’s be honest. Setup has always been where the real admin magic happens. But it can be surprisingly complicated to answer a simple question, like, “Why can’t my user edit this record?” As Cheryl Feldman explains, it’s because Setup has slowly grown in complexity over time. As new products and features were added to Salesforce, Setup pages multiplied. And while that level of complexity will always be necessary, it doesn’t need to be so hard for admins to navigate. That’s where Setup with Agentforce comes in. Whether it’s troubleshooting permissions, understanding formulas, or finding the right configuration page, you can do it all with conversational AI instead of wading through 1,300 pages. How to try Setup with Agentforce right now Setup with Agentforce is currently available in open beta, so you can give it a try in production, a sandbox, or a developer org. Cheryl and her team would love to get your feedback. “We want to hear from admins,” she says, “Does it help you? And what do you want us to solve next?” Right now, Setup with Agentforce can handle common use cases, like user access, formulas, data model, and Flow. And for questions it can’t cover, the agent will pull answers from Salesforce’s help and training documentation and point you to the correct Setup pages to get the job done. A simpler, agent-driven future for Setup Right now, Setup with Agentforce focuses on high-impact, everyday Setup tasks. But Cheryl and her team have big plans for the future. One area they're looking at is multi-step orchestration, which will give admins the ability to create objects or fields, grant users access to them, and add them to a layout in one guided process. Instead of jumping between pages, you’ll be able to complete related tasks in a single conversational experience. Most importantly, we need your feedback to shape what comes next. Jump on the open beta, reach out to Cheryl and her team, and let’s make some admin magic. Check out the full episode for more from Cheryl Feldman. And remember to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Trailblazer Community Group: Salesforce Admins Podcast Episode: Salesforce Admins Blog: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Admin Trailblazers Group Social Love our podcasts? Subscribe today or ! Full show transcript Mike Gerholdt: This week on the Salesforce Admins podcast, well, let's be honest, setup has always been where the real admin magic happens. On this episode, I am talking to Cheryl Feldman, Product Management Senior Director at Salesforce, and longtime admin advocate to unpack how Agentforce is reshaping the way we build, troubleshoot, and think through configuration. From solving tricky user access questions to helping us with formulas and beyond, this is about making setup smarter and more intuitive. With that, let's get Cheryl on the podcast. So Cheryl, welcome to the podcast. Cheryl Feldman: Thanks so much, Mike. I'm excited to be back. Mike Gerholdt : I'm excited to have you back. You were a rockstar at Dreamforce this year, and you're probably going to be a rockstar this whole year. I know in the admin keynote, we demoed Agentforce for Setup. I think I'm using that term right, right? Cheryl Feldman: Setup with Agentforce, but thank you. Yes. Mike Gerholdt : Okay. See, I'm always backwards. Cheryl Feldman: Yes. I was so excited to see that work in the keynote and to see the exciting admin response and excited that we went to beta today. Mike Gerholdt : Oh my God. I feel like I feel as much joy for this as I do when we could customize the homepage. Cheryl Feldman: Yes. Mike Gerholdt : It sounds so small. We're getting in super knee-deep. Let's zoom out for a second because I'm sure there's a lot of new admins that are like, "What are they talking about?" Cheryl, for people that haven't met you and heard you in some of the True to the Core, and some of the Q&A sessions where I feel like Parker just calls on you constantly, could you do a little bit of an introduction of what you do and the fun stuff you get to work on at Salesforce? Cheryl Feldman: Sure. My name is Cheryl Feldman, and I'm a Product Manager in platform. Been at Salesforce about four and a half years now. Prior to joining Salesforce, I was actually a customer for just over 18 years in, starting out as an admin, moved into more leadership and management roles, was running a COE at a very large bank before I joined Salesforce. And when I joined Salesforce, I started in the user access area, specifically managing the authorization side of things, and then started to really advocate that we improve admin experiences. That led me to my role now where I act as the product domain lead for Setup with Agentforce, and improving experiences for admins and really addressing a lot of the issues that admins have in setup because speaking of True to the Core, setup has come up pretty much every True to the Core has ever existed. And that's what I do here at Salesforce. Mike Gerholdt : Which is a lot to say the least. I will fan out a little bit because I remember when Cheryl, you were a customer, you always had questions I couldn't answer, which challenged me. And now to see you on the inside, I mean, it's one thing to hang an identity and say admin developer architect, but that kind of core identity that people have that you made your name on of configuring one of the most complex CRMs to meet business needs without deploying code is just so empowering. And I just, for me, on behalf of all the people like me, I'm just so appreciative to have you inside the PM org championing for people that want to be in tech that don't understand or know how to write code. I had to say that early just because- Cheryl Feldman: Thank you. Thank you for saying that. Mike Gerholdt : ... I don't get to talk to you a whole lot. You're a busy person, especially with the new setup stuff. I mean, I remember demoing permission sets and permission set groups and feeling like, "Holy cow, Cheryl just moved a mountain." But you're like, "Yeah, I can move mountains, but what about planets?" And now you're tackling setup. Where did this come from? Cheryl Feldman: This came from back, I think it was June 1st, 2022, one of our incredible Trailblazers, Andrew Russo, hope he doesn't mind me shouting him out, stood up at True to the Core and said, "When are you actually fixing setup? When are you doing anything, something?" And I had been pushing internally that we needed to do something. And I said, "Well, if not me, then who?" And so I started looking at the usage of setup, and something that it didn't necessarily surprise me, but it validated a lot of things that I believe that the areas and user access were the most used areas and setup, which was the area that I was managing at that point from a product standpoint. And I said, "Well, what are the big issues that we need to sell for admins?" And a lot of it was around troubleshooting. So we introduced the user access summaries, and we did a lot of great work in user access, but we didn't really expand beyond that. And so we said, "What would it take to actually fix all of setup?" And trying to actually do that is... Actually, I'll be honest, it would be an impossible task given the state and that there's no framework behind setup. So when Agentforce started to come about, we said, "Well, what if we looked at Agentforce, and what if we could help admins and rethink setup using Agentforce, and revamp things and revamp the experience that way?" And so that's how everything came about. We started with, let's improve some of the pages to then let's see if we can tackle a larger part of setup, and that's what we've been doing. And Agentforce has helped unlock that for us because I'll be honest, the way we were going previously, we probably would've been here for 20 years because there were over almost 1,300 pages in setup, there's a lot. Mike Gerholdt : Wow. Well, and as I come to find out, I mean, there was no real governance of setup. Cheryl Feldman: That's true. That is very true. Mike Gerholdt : And so everybody, when they created a new product would just add it to the setup menu. Cheryl Feldman: Yeah. Mike Gerholdt : Wow. And I happened to think that the solution actually in 2022 maybe was, it had been invented. I'm sure somebody will call me out on when AI was birthed, but it wasn't in our nomenclature for another year or so. Cheryl Feldman: Yep. Mike Gerholdt : And that was actually the solution because I remember thinking, I was at a world tour in Boston and somebody came up and said, "With the innovations coming out around AI, why can't I just send screenshots of my problems or my formulas to AI to fix it?" And I thought, well, you should just solve that on the platform. Cheryl Feldman: Exactly. And we actually do, and so one of the things that Setup with Agentforce can help you with is to troubleshoot your formulas. It doesn't just troubleshoot user access, because that's one of the things we started looking at, what are customers asking in the Trailblazer community? And a lot of it is, I always see Steve Mo and I always say, "We need to help our agent think like Steve Mo helping out all those new admins with their formulas." Mike Gerholdt : Yeah. I think one important thing, and we don't have to get into the details of it, but one important thing is for you, this was, yes, a really cool thing to come out with, but you also dug into, "Well, but how does this help on the back end? How does this help our customer service? How does this help our admins?" And you don't have to share it, but you really found what the percentage of cases and the reason some of the admins were calling into our customer support line. Cheryl Feldman: Yes. This is probably not a shock to a lot of admins out there, but it was very hard if you are not familiar with SOQL to answer the question, "Why can't my user do a thing or why can't my user edit this record?" That was very complicated to figure out. So unless you had a developer working with you who could run SOQL for you, you ended up contacting support. And so we started looking at how could we help admins help themselves in an experience through natural language. And so that is one of the first use cases that we took what setup with the Agentforce was around that troubleshooting. Mike Gerholdt : Wow. That's amazing. So fast-forward, it's January 2026. I almost said '25. I don't know why. It's like the brain didn't move ahead a year. How are we doing? Where are we at with our agentifying setup? Cheryl Feldman: The exciting announcement as of today, or you'll be listening to us in a couple of weeks. It'll be out for a couple of weeks already, but we are now an open beta. We were, announced our pilot at Dreamforce, so we're now in open beta, so customers can go enable this and try it out. We solve use cases and we handle use cases across user access, formulas, data model, flow, a lot of the most common areas and setup. Now, I don't want to say we handle everything, but we do handle a lot, but we want customer feedback. We want to hear from admins is, does it help you and what do you want us to solve next? And so that's where we are right now. Mike Gerholdt : Okay. And just to clarify, because sometimes I know for a long time as an admin, I wasn't able to do some of the things. Pilot was totally, you had to sign up, and Cheryl had to interview you, and you had to shake hands and there was a vetting process, right? Cheryl Feldman: Yeah. Mike Gerholdt : Open beta for an admin sitting there listening to the podcast means they can access it in their sandbox or? Cheryl Feldman: They can access it in production, or a sandbox, or in a developer org. It just has to be for developer org. It has to be the specific developer org that has access to Agentforce and Data Cloud. Mike Gerholdt : Okay. Cheryl Feldman: And then they can enable it. And so they'll see, if their org has Agentforce and Data Cloud, they'll see a banner at the top of setup that will redirect them to enable this. Mike Gerholdt : Awesome. You're super close to the product. If you were an admin now, what would be the first thing you would suggest an admin should do when they try this out? Cheryl Feldman: I would say think about an issue that you had, whether it was an issue writing a formula or an issue trying to figure out what a user of access was, and ask that question to Setup with Agentforce and see where you get. One of the things that I think is really cool that I would've loved this when I was a customer, because I had to answer this question all the time. You could actually ask the agent, "Show me a list of users that have access to customized application or modify all data or manage users or managing public list fees," and it'll show you a list. There is no great way in the platform to actually figure that out without running a SOQL query. Mike Gerholdt : Right. Cheryl Feldman: And so that is something I'm actually really, really, really excited about. That's where I would start is start thinking about some of the questions that you have. One of the other features that we have is we actually have and help a navigation action. If we don't actually have actions that actually help you do something or analyze something, it will actually take our help in training documentation and generate an answer for you and also help you navigate to the right setup page. Let's say you ask the agent, "How do I setup Agentforce for sales?" Instead of having to go out to the help and training site or out to Google to figure that out, you could ask that right and setup and this will help generate an answer. And then if there's something where you need to redirect to a setup page, it'll give you a link that'll just navigate you to the right setup page. Mike Gerholdt : Wow, okay. I mean, this takes setup from an area where I'm navigating a million little dropdown carets to a whole experience where I can actually query. I remember a while back, it was like we used to say you could talk to your data. Now, you can talk to your setup menu. Cheryl Feldman: Yes. Now you can talk to setup. Mike Gerholdt : In the lifespan of Salesforce products, you said it went from pilot to now open beta. I'm not going to hit you up for timelines, but what is the next step? Cheryl Feldman: The next step is we want to start enabling some more use cases and handling more of setup. We can, I'm sure, share a link to a presentation we did at Dreamforce called the Feature of Setup. We talked about introducing essentially another framework that essentially runs Setup with Agentforce. We're working towards that. And so we handle, we actually started with, we looked at the most used areas in setup and we started with that, but that essentially the most used areas in setup are essentially sit within 40 pages within setup. I mentioned there are 1,300. The next step is how do we expand this beyond these most used common use cases to everything in setup? And that is the framework that we're building. The other thing that we're working towards is how do we handle what we call multistep orchestration and being able to do multiple steps as far as setups. So think about when you create objects or fields, being able to grant access, adding it to layouts or dynamic forms, those are all multiple steps and being able to do those together. That's essentially what we're working towards from the framework. I can't give out exact timelines yet, but those- Mike Gerholdt : No, I don't expect you to. Cheryl Feldman: ... that we're working towards right now to handle more of setup. And so that's what you're going to see start coming as we move throughout the year. Mike Gerholdt : So you envision a world where on Monday an admin could sit, take some requirements, vet out a process, maybe start talking to... I say talking, hello? Typing. Typing to Agentforce for setup and start building an app and then maybe on Wednesday or Thursday, continue building it and possibly even deploy it on Friday and doing it all conversationally. Cheryl Feldman: I think eventually, yes, we will get there. We're handling right now. We consider that. We're starting to think about our building use cases, and I'm actually working, I think you've met my colleague, Tian, and he's working towards the building use cases. And so he would actually be a great person to have on this podcast. Mike Gerholdt : I know. He's been in the admin keynote. Cheryl Feldman: Yes. And so I think we're looking at, I would say overall, yes, but we're not handling those type of use cases yet as far as building an end-to-end app. I would say we're looking at more of the everyday configuration use cases and setup use cases right now, but we're starting to think about right now. So I would say Tian would be a great person to have on this podcast. Mike Gerholdt : Yeah, absolutely. Talk the building part. One thing that I've thought of, and it's come up in other AI conversations is Josh made a mention of it in a previous podcast too of AI almost feels like that first time you get to use a calculator in math class. And I remember my math teacher saying, "Well, the calculator's answer shouldn't be a surprise. You should know what its output is expecting." Two years, one year, I don't know the timeline down the line, it's feasible that an admin could sit down and just conversationally build an app. What are the thought process behind also adding to that so that if they didn't use Agentforce for setup to build that app, that they would know how to jump into the configuration of it? Cheryl Feldman: That's a good question that I'm not sure I have an answer for right now, but I do see that we probably want to do something like that where, because I don't necessarily see a day where setup is going to completely go away. I think it'll still be there. And I think it's important for admins to understand what the agent is helping them do and why. Something we've talked about long-term, and this is further out, is something that I like to call architect in a box. Mike Gerholdt : Ooh. Cheryl Feldman: And I think there's a very big bridge between an admin and an architect. And this is not to downplay the role of an architect, super important role, but I think if you talk, and I played an architect role before I joined Salesforce, so I think it's very much how do we help admins think more like architects on their day-to-day tasks. And so how do we explain to them if they say, "Okay, let me update. If they tell an agent or even go into setup, let me update my sharing model." Doing that on Tuesday at three o'clock is probably not a great idea, right? And helping you think through the changes, what is the impact of the change that you're making and that's where we want to go directionally. Mike Gerholdt : Oh, I like that. I mean, that's being context aware of, you asked me this, but it's say January 29th and our quarter closes in three days, refactoring a sharing model or refactoring permissions might leave salespeople unable to update an important opportunity. What if we do this...
/episode/index/show/buttonclickadmin2/id/40397340
info_outline
How Does Agent Script Give Admins More Control?
03/05/2026
How Does Agent Script Give Admins More Control?
Today on the Salesforce Admins Podcast, we talk to Joshua Birk, Senior Director of Admin Evangelism at Salesforce. Join us as we chat about how to get started with Agent Script and how it helps admins build better AI agents. You should for the full episode, but here are a few takeaways from our conversation with Joshua Birk. What is Agent Script? Big changes are coming to how admins build AI agents with Salesforce, and so I’ve brought Josh Birk on the pod to help us understand what’s going on with the new Agentforce Builder. There are backend updates to the Atlas Reasoning Engine, but Agent Script is where the rubber meets the road for admins. It’s a simple, high-level scripting language that lets you create complex instructions and actions for your agent to follow. Agent Script puts the human in the loop for all the little decisions an AI makes, helping you build more predictable and reliable agentic solutions. If you’re comfortable with coding, you can pop open the hood with Script view, but there’s also a visual Canvas view to help you see how things fit together and make the changes you want. And as always, you can just tell the Agentforce assistant what you want your agent to be able to do, and it’ll list out suggestions for you to accept or decline. Why can’t I copy my agents into the new Agent Builder? The biggest question most admins have about the new Agent Builder is why they can’t just copy and paste their old agents into the new system. The answer is that the new version uses a completely different engine: hybrid reasoning. Josh uses the example of the transition to Lightning Web Components from Aura and Visualforce. LWC was created to align with modern web standards that simply didn’t exist when these older frameworks were developed. The same principle applies to the new Agentforce Builder and the addition of Agent Script. The control that it gives you is a result of the hybrid reasoning engine it’s built around. Don’t panic, learn Agent Script Josh’s biggest piece of advice is to take things slowly. The agents you’ve built in classic will still work fine, and just like with Visualforce, they’re not going to be sunsetted anytime soon. “Salesforce does not like taking away people's toys,” Josh says, “you don’t need to be in a rush to go and recreate your existing agents in the new Agentforce Builder.” Instead, try building new agents with Agent Script. It’ll give you a better idea of just how much more control and reliability is possible. And once you’ve gotten your feet wet, you’ll have a better idea of what you’ll get when you update your existing agents. As Josh says, “Don’t panic, learn Agent Script.” There’s more great stuff from my conversation with Josh about Agent Script and the new Agentforce Builder, so be sure to take a listen. And make sure you’re subscribed to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Blog: Admin Trailblazers Group Social Love our podcasts? Subscribe today or Full show transcript Mike Gerholdt: This week on the Salesforce Admins Podcast, if you've ever been wondering why you can't just copy and paste your old agents into the new version, this episode is for you. I'm joined by Josh Birk, who recently walked our admin relations team through what's really changing with Agent Builder and Agentforce. And more importantly, why it matters. We're going to unpack a little bit of hybrid reasoning, what determinism really means for admins, and how Agent Script gives you even more control without losing all of that magic. So with that, let's get Josh on the podcast. So Josh, welcome back to the podcast. Josh Birk: Thanks for having me, Mike. Mike Gerholdt: Well, we know that a few people are vibing now because of the dulcet tones of Josh Birk. So in all honesty, you were in a team meeting last week and you were kind of updating the team on stuff that's coming out for new products and beta products. And the biggest thing that you always explain is like, so why is this important? Josh Birk: Yeah. Mike Gerholdt: And of course, every now and then you end your little skit with, "Thanks for attending my TED Talk." Josh Birk: Yeah. Mike Gerholdt: And I remember last week's meeting, I was like, "Oh man, that's a podcast. I need to do that." Josh Birk: Yeah. Mike Gerholdt: Because the way that you would explained what was new with Agent Builder and Agentforce, I was just like, "Oh, that totally makes sense to me." So this is where we're at and this is what I was hoping we would talk about. Josh Birk: Yeah. And I'll have to try because I kind of remember getting into a very weird, almost like conspiracy theory level stream of consciousness. Mike Gerholdt: Oh, okay. Well, it didn't feel that way. Josh Birk: Good. I think that's years of experience of thinking in streams of consciousnesses, whatever the plural of that is. So yeah, so I think to kind of recap a little bit, so there were two things that I think are kind of important to consider about the new Agent Builder. One is what we mean by hybrid reasoning and what we mean by determinism. And then the second thing is, where does Agent Script fit into that? So the question we were kind of tackling was, why didn't we just go back and revise the old builder? Why can't we just reiterate on it? Why are we starting completely new? Mike Gerholdt: And I think you even said, not to interrupt you, but to interrupt you, why can't I just move my stuff over? Why do I have to rebuild it? Josh Birk: And why can't I just move my stuff over? Why isn't it just copy and paste? And Agent Script is part of that. So the original engine wasn't hybrid, is I think the easiest way to put it. So this came with a conversation with product management where I was trying to describe the differences between the old builder and the new builder. And they're like, well, let's ... The old builder wasn't doing anything wrong, so much is that the new builder's doing something more. And what the new builder's doing that is more is that when we say hybrid, when we think of AI, we think of AI being autonomous, making its own decisions. So we had the Atlas reasoning engine that's like, "Oh, Mike is looking to find a description on a certain account. I'm going to take these actions. I'm going to put them together into a plan and then I'm going to try to make that work." What we didn't have was the human in the loop portion of that. And where Agent Script really comes into play is it allows users to basically be like, "In this scenario, I want you to behave like this." So when we say hybrid, we mean, yeah, we're letting the engine kind of do its own magic. But then we're also applying almost ... I don't want to say flow because that's such a loaded term for a Salesforce community. Mike Gerholdt: Yeah. Josh Birk: But well, let's go to the name, right? A script, right? A script that a human can actually control personality, control tone, control behavior. When you see this, do this, when you need to call it human. So kind of giving this very human-friendly dialogue that you can then apply to the agent. And so that's how we provide more determinism. So the reason you can't just copy and paste your old agent into the new agent is the old agent doesn't have that connection to Agent Script. And so it's literally missing a piece of the puzzle to run the engine kind of thing. Mike Gerholdt: Right. No, I think ... I mean, hearing you explain it, sometimes, at least as an admin, I was always like, "Oh, so they just came out with another version because they had to stick another layer of something on." Josh Birk: Yeah. Mike Gerholdt: Or like a car analogy, they changed the fuzzy dice. But this is, no, no, no, no. We're really moving from this version, which ran a certain way to an even better version. And that's the part that I feel like ... Because you brought up Flow, I feel like we've been through this with Flow. Josh Birk: Yeah, totally. Mike Gerholdt: We got like a business process ... Business management process is what it used to be called. And then that thing went through iterations and then we found a different engine to run it on. And then we had ProcessBuilder for a while and then now we have the new Flow. And it's like sometimes it's not just the interface that changed, but it's actually the guts behind it. Because I mean, I can't recall ... It's sometime between after the pandemic and today that AI came around. And I feel like that's the window that works in my brain because I can't put down a time. Josh Birk: Totally. Mike Gerholdt: Was it 2022 or is it 2023? I don't know. Post pandemic, before today, AI happened. But just when you thought, okay, I've kind of got this thing figured out, nope, all of this stuff changes. Josh Birk: Yeah. And it happens in technology, right? It happened with the Lightning Web Components. We really kind of had to go back and take the new ECMAScript format and take the new ECMAScript standard and really rebuild the way our web components work so that they work in a standard based way. So that they're more similar to React and more similar to these other frameworks. And there wasn't really a way to go back to Aura, now it's called, and just be like, "Hey, Aura, be more standards based." Mike Gerholdt: Right. Josh Birk: Because Aura was effectively just ... It was pre-standards is the easiest way to put it. So this happens, right? Now, if people are listening to this and like, "Oh gosh, I have all this work in my old Agent Builder, what am I going to do?" First of all, Salesforce does not like taking away people's toys. Mike Gerholdt: Right. We've learned that. Josh Birk: We've learned that. I've said this anecdote in workshops so many times because when we did release Lightning, one of the most common questions I got was, "Are you going to end of life Visualforce?" So I went to the product manager, might have been Skip at the time. And I'm like, "Hey, when would you end of life Visualforce?" And his response was, "When I pull up a dashboard and I see there are zero active users using Visualforce, and then I'd probably wait three months after that." Mike Gerholdt: Just in case. Josh Birk: Just in case, right? Mike Gerholdt: Yeah. Josh Birk: So the classic one is not going anywhere anytime soon. But I think the other important thing is a lot of our messaging when it comes to building out agents is that you're already an Agent Builder. You're already somebody who can build AIs because you know Flow, because you know Salesforce. Now you've learned topics, now you've learned actions and all of that kind of stuff. All of that is still true. It's all still true. And so you're still getting to use your existing skills in order to control it. It's really just adding the skill of knowing how to write out the Agent Script. Now, one of the cool, really cool things about ... Well, there's two cool things about the new builder that I think will really help people with that. One is the Canvas mode. And so Canvas mode takes what you write in Agent Script and puts it into a visual UI interpretation. And so this is very handy for two things. Number one, it gives you a little bit of kind of a predictive window as to what your script is going to be doing. It also gives you a very visual way of thinking about it. It's also a really neat way to go look at somebody else's Agent Script and translate it. So you don't have to read through all the thing. You can flip over to Canvas mode and be like, "Oh, that's trying to get these parameters in order to do these actions." The other thing going back to vibes and this new AI coding AI is that you'll have an AI assistant in Builder to help you build the AI. So you're not alone. Agent Builder itself will also help you create these things out. So my biggest statement back in the Lightning web component today was don't panic and learn lighting. And now it's, don't panic and learn Agent Script and you're going to build out some really powerful solutions. Mike Gerholdt: Well, and I think back to ... I mean, I was chatting with Jennifer today too, and just the way that we started using AI. And I remember, I mean, not that long ago doing workshops and saying, "Okay, you need to tell the AI your role, its role. And then what you want it to do and what it can do and what it can't do." And for everything, whatever model you were using, I remember writing prompts a certain way. And I feel like for this and with Agentforce, especially for admins, this is the least set it and forget it kind of thing. Josh Birk: Yeah. Mike Gerholdt: Because you can always go back and make it better. Josh Birk: And tweak it. Yeah. Mike Gerholdt: And this is- Josh Birk: And tweak it in theory. Yeah, exactly. In very small corners too. Mike Gerholdt: Small corners that make a huge difference because like with Flow, I mean, half the time I'm just excited that I got it to do what I needed it to do. And then I look at what Jennifer built and I feel like I'm over here rubbing two sticks together. Josh Birk: Together. Mike Gerholdt: But with Flow, that's your goal, right? You just want that, okay, boom, it's done. But with agents, people are interacting with them, they're asking them questions, I can make this so much better and I can make this better. Here's the clue when people stop using them. To your point of when you would replace Visualforce. Well, if you're looking at your Agentforce dashboard and seeing, "Hey, your users aren't using it." well, now might be a time to like, "Well, let's switch over, let's round some edges, let's change some things." And I would say this is like an every two weeks kind of thing. Which sounds crazy, but the amount at which you're going to find feedback from your users. And the amount of rounding of edges, I worked myself into an English problem here. It's going to be like, you're going to find like a star and then it's going to look like a ribbon of rounded edges that you have to work on. Because as you use that new builder, now your users are happier and being more proficient. And I think like even today, when Jennifer and I were talking, we were talking about building Gems in Gemini. And I said, "One thing that I do is I'll actually have another window open and ask it to refine my instructions in Gemini." Josh Birk: Oh, yeah. Mike Gerholdt: Because the way that I write the instructions works for me, but maybe not for it and it can refine it. And I've seen it take thousand word prompts and reduce it down to 200 and it's actually faster. And so having all of that is sometimes there's a lot to keep up with. But it's all for the better because now you're like click click, oh wow. And then you get that ... I think it's always, at the end of the day, the admin is going after that first look on a user's face and they're like, "Ooh, this- Josh Birk: Nailed it. Mike Gerholdt: ... just made my life so much easier." Josh Birk: Exactly. Yeah. And I think to kind of append to that, again, not to overuse this analogy of Visualforce. But it's like don't be in a rush to go and recreate your existing agents in the new Agent Builder. If you have a new agent you're building, I would start with the new Agent Builder and go from there. Two reasons, don't fix what's not broken. And also you'll learn more about Agent Script and things like that. So when you do decide to move your old agent to the new builder, you'll be more educated on how to do it. Now, huge forward-looking statement, asterisk, safe harbor, there is a conversation internally about creating a porting tool of some kind. It's on a roadmap somewhere, no guarantees. Mike Gerholdt: Oh. Josh Birk: But I do think people should know where we are looking at that kind of thing. So stay tuned because we do want to make this as easy as possible. Mike Gerholdt: Yeah. Yeah. But also there's some joy in rebuilding things and maybe that's just me. But I've always- Josh Birk: Seems like part of the tech world. Yeah, exactly. Mike Gerholdt: Yeah. I mean, there's that and there's also, it reintroduces you to things that at one time you were good at and brings it back. And then I think too, anytime that these new features come out, one part is, okay, well, here's the tech to change it. But also, has the process changed? Has the stuff that people do internally changed? And that might also be true. And the other point you brought up, which was really salient, just because you can change it, doesn't mean you should right now. Josh Birk: Yeah. Mike Gerholdt: One thing that ... And I was a part of this, doing a big org rollout. What are the other changes that your users are experiencing within your organization? And so you adding to that, is that one layer too many? Or when would be a better time? Like we always joke, like you're not going to roll out a sales opportunity change the last week of the quarter. Josh Birk: Right, right. Mike Gerholdt: So picking your timing, this means that gives you time to test. Josh Birk: Yes. Yeah. And I think this goes into a lot of the conversation we're having about admins in the AgentiCare. Like what kind of a steward are you being? And how are you using those existing instincts and skills as an admin to not topple over the apple cart, but now you're doing with AI? Mike Gerholdt: Right. Because in my opinion, admins are getting paid for their judgments. Josh Birk: Yeah. Yeah. Agreed. Mike Gerholdt: Josh, thanks for coming on and helping us understand the new Agent Builder. Josh Birk: Thanks for listening to my TED Talk. Yeah. Mike Gerholdt: I knew you had to fit that in. And we've got just enough time that Daryl can hopefully get his boat from home to the internal ... It's not internal. Inside the podcast joke. Josh Birk: Inside the podcast joke. Mike Gerholdt: You have to go back to the Daryl Moon episode- Josh Birk: Yeah, exactly. Mike Gerholdt: ... and then it all makes sense. It's the key to unlocking the mystery of the Dan Brown episodes. I don't think so. Josh Birk: I love it. Mike Gerholdt: Anyway. Josh Birk: I love it. Mike Gerholdt: Have a good day, Josh. Thanks for coming on. Josh Birk: Cheers, my man. You too. Mike Gerholdt: Huge thanks to Josh for coming on very impromptu and turning what could have been, I think, a little confusing of an update into something that actually makes sense. New engine, new script layer, AI helping you build AI. Awesome. I like it. So remember, don't panic, learn Agent Script, and it's all about having fun. And like any powerful tool in Salesforce, it's less about flipping a switch and more about steady improvement over time. So if this episode helped you clarify your next steps, share it with another Salesforce admin who's navigating some AI right now. And until next time, we'll see you in the cloud.
/episode/index/show/buttonclickadmin2/id/40314380
info_outline
What Are True to the Core Deep Dives at TDX?
02/26/2026
What Are True to the Core Deep Dives at TDX?
Today on the Salesforce Admins Podcast, we talk to Kate Lessard, Lead Admin Evangelist at Salesforce. Join us as we chat about the new True to the Core Deep Dive sessions coming to TDX. You should for the full episode, but here are a few takeaways from our conversation with Kate Lessard. True to the Core Deep Dives are coming to TDX If you’ve been enjoying the True to the Core Deep Dive video series, we’ve got something exciting coming up for you at TDX. Each month, we host a virtual Q&A session with Salesforce product leaders about a topic chosen by the community. It’s a great opportunity to dig into tough use cases and thorny questions around a specific area of the platform. The video series has been so helpful that we’re going to take it a step further. This year at TDX, we’ll have four live True to the Core Deep Dive sessions on specific topics, like flow and automation. These will be can’t-miss events, and Kate Lessard is here to tell us what she has in store. Direct access to product leaders “The chance to actually connect directly with the Salesforce product leaders on those core platform topics is worth its weight in gold,” Kate says. It’s not just about solving problems; it’s about understanding the thinking behind why things work the way they do and where the product roadmap might be headed. If you’re newer to the Salesforce ecosystem and worried that things might get too technical, Kate encourages you to check out these breakout sessions. “There's a lot that you can gain from just being in the room, seeing who those product leaders are, hearing what's top of mind for them, and how they're answering questions,” she says. You can better prepare for what’s coming next on the platform, and you never know when a random use case you came across might apply to you and your org. And even if you can’t be there in person, you can participate virtually on Salesforce+. How admins should prepare for TDX TDX is such a great opportunity to level up your technical skills and get to know the product better. One thing I used to do to prepare was conduct an audit of my Salesforce org. What products are we using? Where are we having problems? Where are we going next? Figuring out what you’re interested in can help you make the most of your time at TDX. Hopefully, you’ll come back armed with information and insights to help your stakeholders chart a path forward with Salesforce. Listen to the full episode for more from my conversation with Kate about True to the Core Deep Dives at TDX, and be sure to catch her on Kate Clicks Through It. And don’t forget to subscribe to the Salesforce Admins Podcast. Podcast swag Learn more Salesforce Admins Podcast Episode: Saleforce Admins Blog: Saleforce Admins Blog: Saleforce Admins Blog: YouTube Series: YouTube Series: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: Welcome to Salesforce Admins podcast. This week I'm joined by Kate Lessard to talk about the brand new True to the Core Deep Dive sessions coming to TDX and why they might just be your favorite part of the event. We dig into how these sessions are shaped by community feedback, what Salesforce admins, especially the newer ones, can expect, and how to ask better questions to get some real value from our product leaders. Kate also shares some practical tips on preparing for TDX, so you walk away ahead of the curve. So with that, let's get Kate on the podcast. So Kate, welcome to the podcast. Kate Lessard: Thanks. Happy to be back, it's been a while. Mike Gerholdt: It's been a while, yes or no, except if people are attuned to our YouTube page. You're on Shorts whenever I'm not publishing something about the podcast, so they're staying up to date. It's not like you've gone missing. Kate Lessard: No, no. I love being able to just share those top of minds and the things that are important or the lessons that we're learning as we really dive into Agentforce and all of the different things that we're learning over here, especially we just went through a new release and there are always fun new features to call out. Mike Gerholdt: Constantly. And we're heading into spring and spring means, besides new release, TDX. Kate Lessard: Yeah. That's an exciting time of year. Mike Gerholdt: I think so because it's kind of our shake the winter off, go out, see some fresh demos and some new stuff. I know as we're recording this podcast, kind of content is in flight and things are being worked on and ironed on, and so is it fair to say sneak peek or I forget, we use all these terms and sneak peek means something, but you have stuff in the works for TDX that you're working on that we're going to talk about. Kate Lessard: Yeah, so I am working on deep dives, which is really exciting. So you might have been following along and watching our recent True to the Core Deep Dive virtual series, and we are bringing True to the Core Deep Dives to TDX as well, so that is what I have been working on lately. Mike Gerholdt: Wow. So tell us more about deep dive. If somebody hasn't seen any video or anything about deep dives, what are they? Kate Lessard: So deep dives are really focused on diving deeper into community driven core Salesforce platform topics. So the virtual sessions, our community actually provides feedback and guides what those topics are. So they've been chosen using input from things like the idea exchange, past in-person True to the Core keynote sessions, as well as our virtual True to the Core sessions and feedback that the community shares. So it really gives the community a chance to guide the conversation and ask the questions that are really important to them. Mike Gerholdt: Yeah, and you said True the core, which is always a popular session for anybody attending our events. Kate Lessard: And the thing about that True to the Core keynote, to just hop in there. Mike Gerholdt: Yeah, no, no, no. Kate Lessard: It is, it's really popular. I remember as an admin, it was one that I never wanted to miss. And whether I had a question to ask or whether I wanted to just sit back and listen and hear what other people were asking and get ideas and cheer with them on the different wins and things that we got approved and that we saw progress on, it's so short that time flies by and inevitably there's just not enough time. You always have people that are still lined up and waiting to ask questions. And then we also have people that are submitting questions online, so there's no way to get to all of them, and I feel like these deep dives give us a chance to actually continue the conversation about certain large topics, things like flow and automation that people can come in and they can specifically ask their flow and automation questions that don't get covered during the keynote because there's just not enough time. Mike Gerholdt: Yeah. Well, that's the beauty of like the times I've been at user groups where you sit down and somebody says, "We're trying to do X, Y, and Z." And you go around the table like a bird's of a feather and you hear three different ways to do it and that person's face is like, "Ugh." And you kind of have to reassure them like there's three different ways to do it, all three are right, but the one that's right for you is the most right, which is why I think the True to the Core stuff is always so interesting because you kind of get people coming in with those edge cases. Kate Lessard: Absolutely. And the chance to actually connect directly with the Salesforce product leaders on those core platform topics, that's worth it's weight and gold I feel like because our community can get clarity on both upcoming features and some of the thinking behind the roadmap decisions and feel much more connected to the product and like they're participating and influencing these future releases by asking these questions and by sharing feedback, it's really important and impactful. Mike Gerholdt: Yeah, and that's the ... Our product leaders are presenting at a lot of our events, but TDX, I know having worked on the track, they are really heavy into all of the product people. So the chances of you going to a session on flow or on Agent Builder or Lightning Web components, the chances of you seeing the product manager present and talk about that is really high at this event, moreso than a world tour or, well, Dreamforce, there's just a problem of you need like 100 of you to go to all the sessions. Kate Lessard: Yeah, absolutely. And to run across all of San Francisco it feels like to get to all the sessions on your agenda. Mike Gerholdt: You put the Dreamforce marathon on your step tracker. Kate Lessard: Yep. Wear your running shoes, you're going to need them. Mike Gerholdt: So True to the Core started a long, long time ago, way back when Salesforce launched Chatter, which was like, Hey, are you paying attention to the core features? Let's break this down. If you're a new Salesforce admin and you're like, "So do I go to this True to the Core Deep Dive?" I could foresee somebody sitting there and being like, "I don't understand why they're cheering, why we get email updates when flows error out or like a obscure feature." If you were a new Salesforce admin and you're going to the True to the Core Deep Dive session, what should you expect? Kate Lessard: If you're a new admin and you're kind of hesitant to go, I would encourage you to just go check it out and sit there and listen. Even if you don't have a specific question that you want to ask, I think that there's a lot that you can gain from just being in the room, seeing who those product leaders are, hearing what's top of mind for them and how they're answering questions because they're going to get questions about how they've built things and what's coming next and being able to have that insight so that you can start to think, "Oh, this is how I do this now and maybe I'm new to this, but this is what is being developed, this is what is being worked on, this is what's on the roadmap, this is kind of the future state." I feel like it helps you become a better admin and prepare for what's coming next as well as to get some really insight into those use cases, as you mentioned, at community groups that get discussed so frequently because people will come with specific use cases and things that they've run into and really specific questions as well. And I just think it's a ton of information to take in and just being in the room is really helpful as a new admin. Mike Gerholdt: I couldn't agree more. Let's talk about, you said questions about 100 times. Kate Lessard: Yeah. Mike Gerholdt: Well, I point it out because I've done a lot of webinars, this is going to date me, Jesus, back in the AM/FM radio days. What's that? But I've done a lot of webinars, a lot of live events, like release [inaudible 00:08:39] live, a lot of Q&A stuff. I've even done Q@A's after presentations. You've done a lot with this True to the Core content. For people going to TDX or people asking questions at these events, what is your advice on how to get the most out of it for just yourself and for everybody in the room? Kate Lessard: That is a great question and there I go using the question word again. Mike Gerholdt: That's okay. Kate Lessard: I think that how you ask a question is really important, and we do, even as we do the True to the Core Deep Dive virtual series, we even cover how to ask a good question and context is going to be key. If it's a specific use case, provide as much information and context as possible. If you know what specific feature or even which product manager you have a question for, being able to call them out and giving that information, super helpful. If not, we'll get there. We do have our product leaders that will be on stage and they'll pull in the different product managers that might be able to answer that question for you. But I think that context and just having a well-thought-out use case with all of the information that you want to bring into the question is going to be the most important thing. Mike Gerholdt: Do you often run into, because I would sit in these sessions a lot and be like, "Oh, I totally want to ask a question, but everybody's going to think it's silly"? Kate Lessard: Yeah, no silly questions, right? I think that that's always a concern for all of us. We want to be seen as subject matter experts in front of our peers. But if you have a question, even if you think it's silly, like nine times out of 10 somebody else has that question, and maybe it's even something that one of those newer admins in the room that's just there to listen and gain exposure, they have no idea. And you're going to be bringing that to light and getting answers for yourself and for a number of other people. So I always say that there are no questions you shouldn't ask, and there are different ways that you can ask them as well. I think if you're in the room, you can get up and ask the questions at the mic. We're going to have a way to submit questions online, and that way we'll also be able to bring in community members who aren't in person. So this will be on Salesforce+ for people to watch as well. So I think that just deciding what way you're comfortable asking the questions is just going to be important for you, but get the answers that you need. Mike Gerholdt: Yeah, I think back to some of the True to the Cores that I would attend as a customer, and sometimes it's really making friends with the people sitting around you and talking with them because there have been times when just happenstance I had an issue and the person sitting next to me is like, "Oh, I've run into that too. You know who you need to talk to?" And then they'll tell you and it actually saves you from getting up and asking a question and moving to another question. And then you found somebody in the audience that you can be a session buddy with, so I think that's super important. Kate, when you were heading to a session like this as a customer, what would some of the homework be? Because admins have about, oh my, 40, is it 45 days to the TDX? I think it's about 45 days. Kate Lessard: Yeah, it's coming quick. Mike Gerholdt: What should they be doing for homework to get ready for sessions like this? Kate Lessard: Well, we've never had a session quite like this before, so that's kind of a trick question. But I would say for all sessions in general, get on to the sessions on the TDX website and pick out the ones that are going to be most impactful to you. There's just so much there, it's impossible to attend and do them all. So pick the ones where you have questions, where you need more information, where you really want the chance to communicate directly and hear from those product leaders because as you called out, that really is the thing at TDX that makes it different, that makes it special. The opportunity to actually have those conversations with the product managers, with the product leaders and to get to know the product better, it really ups your technical skills. Mike Gerholdt: I couldn't agree more. I always feel like I was doing somewhat of a technical audit before I'd head to an event, just to know what I had on my company's roadmap and some of the features that maybe I was running into that I had issues with to come back, and it's always nice to check those off and take notes. Kate Lessard: Yeah, that technical audit, that's a really great point that I think that people should consider. Do an audit of your system. What are you using? What are you having issues with? Talk to your users about things that they might have questions about or struggles they might have. And really talk to your stakeholders as well about what their vision is for the future. Because you might be able to go to sessions and get information about products that you don't have but that you're considering and being able to come back and provide that information and a recommendation of either, "Yes, we should move forward," or, "This is the implementation plan," or, "This is the guidance that I got on this," I think that that is going to be hugely impactful for your organization. Mike Gerholdt: Oh, especially when they say, "So what did you learn? How much did this event cost us?" And you're like, "Well, I got to talk to these people and holy cow, I'm like four months ahead of the learning curve." Last question, how did you balance in-person sessions versus broadcast sessions and repeat sessions? What was your strategy? Kate Lessard: I think that with these deep dive sessions, the goal was to give people as much opportunity as possible to answer or to get their questions answered and to get that live time. So our goal is to have this after the True to the Core keynote so that if people were waiting in line or had questions that they were unable to ask, that they can come to these deep dive sessions and they can continue the conversation and really get the information that they need. So thinking of that, we really were particular about the topics we chose and made sure that we went to the idea exchange and we went to our survey results we get from the True to the Core Deep Dive sessions, the ones that are done virtually, where people provide feedback about what they want to hear more about or where they have questions. And really just trying to give that space for people to be able to come in and continue the conversation. So because of that, all four of our sessions will be on Salesforce+, and then we've tried to kind of divide what will be able to be live versus video on demand so that we can reach the broadest amount of people. Mike Gerholdt: Makes sense. Kate, you have a series on YouTube. You want to talk about it real quick? Because I feel like it's a really cool thing that, as a follow-up to this podcast, people can click over and watch some fun YouTube videos. Kate Lessard: Absolutely. So I am very much a person who learns by doing, as I think many Salesforce admins are. We want to really get hands-on with the different features that we're learning about, that we're delivering to our organizations and become those subject matter experts. So I have a series on YouTube, it's called Kate Clicks Through It, where we take one of these backend processes and we actually walk through it step by step, and I give you a chance to be in either a developer edition or a Trailhead Playground and actually follow along and click along beside me so that you can get hands-on versus just watching a demo and then trying to go and replicate it. And it's been really fun to do. I get to dive into these different topics that I think are really important for admins. And they can go ahead and do it along with me, learn by doing. And they're pretty short and sweet, they're all under 10 minutes so that our admins can go ahead and click through it with me on a lunch break or if they're in between meetings or if they just want a little bit of fun learning outside of their regular workday. Mike Gerholdt: I like it. It's awesome. Kate, thanks so much for coming on the podcast. Kate Lessard: It is my pleasure. I hope to see everyone. I know I'll see you, but I hope to see all of our awesome admins at TDX and hopefully in these True to the Core Deep Dive sessions. Be sure to add them as favorites when you go and look at your sessions on the TDX site. Mike Gerholdt: Well, I want to thank Kate for coming on and helping us get prepped for the new True to the Core sessions. And I think they're going to give you a chance to really lean in. I think she gave you some really good...
/episode/index/show/buttonclickadmin2/id/40228655
info_outline
How Do I Understand a Complex Salesforce Formula Quickly?
02/19/2026
How Do I Understand a Complex Salesforce Formula Quickly?
Today on the Salesforce Admins Podcast, we talk to Gaurang Mathur, Senior Product Manager on the AI App Development team at Salesforce. Join us as we chat about how Setup with Agentforce is changing the way admins create, modify, and understand formulas. You should for the full episode, but here are a few takeaways from our conversation with Gaurang Mathur. The massive scale of Salesforce Formulas If you’re like me, you’ve spent more time than you’d like combing through your formulas for missing parentheses and commas. That’s where this week’s guest, Gaurang Mathur, comes in. As a Senior Product Manger on the AI App Development team at Salesforce, he’s trying to make formulas a little easier for admins with Setup with Agentforce. According to Gaurang, there are 5 billion formulas executed on Salesforce each month. They’re created, managed, and modified by 7 million admins and developers. The scale of formulas is truly massive, and they’re a fundamental building block of Salesforce. However, that means they’re difficult to work with, and that’s where the Setup with Agentforce beta comes in. How Setup with Agentforce simplifies formulas We know that AI is really good at summarizing, debugging, and processing large amounts of information very quickly. Setup with Agentforce lets admins harness those agentic powers to make formulas easier to work with. You can diagnose problems, change an automation, or add something new. Gaurang sees Setup with Agentforce as a game-changer for admins working with inherited orgs. If something isn’t working or you need to make changes, you’re usually spending hours wading through formulas to sort out what’s essential and what’s just buried tech debt. With the beta, you’ll be able to get a helping hand from an AI agent that can validate, describe, and even fix your formulas. Instead of having to figure out what does what, you can focus on making decisions and building something that works faster than ever before. What’s next for AI in Salesforce As far as what’s coming next, Gaurang thinks that AI will continue to amplify what we can build in Salesforce. As he points out, we’ve actually been using AI for almost a decade in things like navigation apps, recommendations, and even autocorrect. There’s no reason to be intimidated. Looking forward, his team is looking at ways to use AI to collect feedback and make adjustments that make it even easier to work with and more powerful for users. Make sure to listen to the full episode for more from Gaurang about formulas and Setup with Agentforce. And don’t forget to subscribe to the Salesforce Admins Podcast so you never miss an episode. Podcast swag Learn more Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Salesforce Admins Podcast Episode: Trailblazer Community Group: Salesforce Admins Blog: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Admin Trailblazers Group Social Full show transcript Mike: Welcome to The Salesforce Admins Podcast. Today on the podcast, I'm joined by Gaurang Mathur, product manager on the AI App Dev team at Salesforce, and we're going to talk about something every admin has wrestled with at some point, which is formulas. And we're also going to hit on how setup with Agentforce is changing the way we create, modify, and even understand them. Hooray. I can always use some formula help. So Gaurang is going to share with us how agents can validate, fix, and describe formulas right inside setup, and why that matters, especially if you've inherited an org or you need to do something super fast. We also touch on Agentforce Vibes, which you heard in last week's episode with Tiaan and some developer workflows and kind of what it means to build smarter with tools using AI. So if you're like me and you've ever stared at missing parentheses and wondered why your formula won't save, this episode's for you. And with that, let's get Gaurang on the podcast. So, Gaurang, welcome to the podcast. Gaurang Mathur: Hey, Mike. Thank you. Thanks for having me. Mike: Absolutely. Well, after we talked with Cheryl, it was important that I get a lot of the people that she works with on the podcast, because you guys are doing some really exciting things around set up with Agentforce, I think I said that correctly, and just some of the things with agents. So let's get started with tell everybody what you do at Salesforce and how you got there. Gaurang Mathur: Yeah, sure. Thanks. So my journey has been pretty interesting to where I am today. I started off as a PM for the caching group, which is again, a platform product. Platform Cache is something that's widely used, and that queue is helping a lot of ISVs to accelerate and perform, make their applications better perform. And over the past five years, I've also managed products in the Experience Cloud Search space, and now I'm currently managing two Agentic products, which is again, hardcore platform products meant for our Salesforce admins and developers, the superheroes in the whole ecosystem of Salesforce. Mike: Yeah. Gaurang Mathur: And as part of that, as you mentioned, I closely work with Cheryl on the Agentforce setup, or setup with Agentforce, and over there, I am managing the space for formulas. I think that's where I'm going to talk mostly about in today's podcast. And the other product that I'm managing is Agentforce Vibes, which is again, zero to one product that we have all built over the last few years. There, I'm creating an MCP tool, which is basically making that agent much more smarter and getting a lot of ecosystem teams within Salesforce to contribute that, and then make it better in generating whatever metadata that it generates. So yeah, that's my portfolio here in Salesforce. We are part of this whole group called as AIF Dev. That's where we've got Tiaan, I think one of the other guests in the previous podcast, Cheryl, and Manish. These are the folks that I'm working with, and excited to be in this whole space in the current year. Mike: Yeah, wow. Two products. So you sleep only a couple hours a night, I'm guessing? Gaurang Mathur: Yeah. I mean, these days, agents have been keeping us alive. I mean ... And so yeah, I'm just, as I sometimes joke with my family, I'm making agents smarter. That's what we work on, so yeah. Mike: Right, right. Well, let's talk about formulas because I think I go back to, boy, 10, 15 years ago, a staple in admin track was always a best practices on formulas and validation rules. And I always felt like formulas as an admin was as close as some people felt to getting to code, because we could carry over a lot of the stuff that we'd learned in different spreadsheets into formulas. So tell me a little bit about what you're doing with formulas and setup and how that works. Gaurang Mathur: Yeah, thanks. I mean, I think you touched on a very fundamental point. And before I get into the actuals of what we are trying to do, but I'd like to double click on what you just mentioned, Mike, which is formulas have been a very legacy product within Salesforce. They have been integral in the whole platform since, I mean, a very long time, and I'm quite excited to carry that legendary thing. I'll just share some numbers with you in terms of how big and the scale of impact that formula has in the whole ecosystem. So almost on a monthly basis, there are five billion formulas that get executed across Salesforce orgs, and seven million admins and developers manage, create new formulas, or modify formulas every month. So that scale is fairly huge, and these formula engines and formulas are supported across almost 40 odd Salesforce services, which includes flows, business rule engines, SLAs, fields, validation rules and whatnot. So like I said, the scale is fairly huge, and we've seen admins using it, developers using it, building formulas quite often, but there has been a constant ask and problems that we've seen that admins figure out as is creating them. Somebody is coming up in a new org, and then just trying to modify those formulas. So those scenarios have been quite error-prone and time-consuming processes, and sometimes challenging for admins. So with agents coming in, we asked admins, "How do you see agents helping you, and what is that agents should do for you?" And among the top 10 asks, managing formulas was right number three and four. So that led us all to make a dedicated decision and team that we should be improving and working towards creating a better formula management experience for our admins with the set up with Agentforce that's coming in. And we did a few releases. I mean, it was one of the few things that we did for formulas, I mean, almost year back. And now with the beta coming in for setup with Agentforce, it's right up front and center. So back to your question then, "What exactly we are doing?" So we are, as part of this beta that's coming up in TDX, we are giving options to create a formula through a set up with Agentforce agent, modify it, fix it, and let it also describe what this formula is. So yeah. Mike: You end very casually, but that's kind of really big. I know, if you were to look back in my search history pre-AI, 90% of it as an admin was, "Why doesn't this formula work?" And I think when AI first came out, that was one of the ... I remember I was at a user group, and that was one of the prominent use cases for using an AI, was to check your formulas, because formulas in Salesforce are much like they are in various spreadsheet programs, so you can kind of understand it. And I remember thinking to myself, "Oh, man, the hours this would've saved me if I could have just copied and pasted a formula and had an AI read it and be like, 'Oh, you forgot a parens, or you missed a comma again,'" you know? Gaurang Mathur: Correct. Exactly. I mean, and those mistakes makes us human, and this is where agents can help and chime in to just call out, "Hey, this is the error, this is the fix," and that's exactly the first principles that we are considering as we build these agents, as we are coming up with this whole agent experience for our admins. So let me just share you one example for that, I mean, what we've just recently demoed to some of the customers and which was very exciting overall. So imagine you are in a company that's managing volunteer experiences and other things. And out there, you have a formula, which is helping you calculate the number of attendees that are there, the volunteers that are going to be participating in that event. And you have all that formula baked in and all that, but suddenly, the admin who was walking on that leaves, or there is change, and now you are on your own and you just have to just kind of get things going, because your event is coming in, or there is some business requirement that comes in. So it gets difficult for anybody to just jump in and get going right off the bat. In this case, we are coming up with modifying a formula, wherein when you see this formula there and you open the agent through the setup page, the formula page, out there, the first thing that the agent does is it's just simply validating the formula and helping you understand what this formula does. Once you get that, now it basically helps you understand and comprehend, "Oh, this is what this formula is, these are the conditions, this is what it'll do." Now, when you have to make some modifications, you can now just simply prompt it, and then the agent does the rest of the job for you in a very quick and easy way. So this is something that's really profound, and while we were sharing it with our early users, that was a welcome, and I'm hoping the community has to see, and they'll listen, try it out, and then figure out what it does. Mike: Yeah, I mean, thinking through, there's two parts to really what you described, which was it can do it, but then, the second part was it can describe, and that's just such a big part of everything that we do. Plus I always feel like, at least for math and formulas felt like math, if I could see a finished product, then I was more apt to kind of understand and work through it as opposed to kind of come to that solution on my own. But you mentioned Agentforce Vibes, and I think that's also, I don't know if you could say companion product, but it's another way to develop on the platform. How have you kind of found some crossover in what you're working on for Agentforce for setup with Agentforce Vibes? Gaurang Mathur: Oh, so you can call them as different facades for different personas and end users. So let me help you with what we are trying to do with Agentforce Vibes. Agentforce Vibes is quite focused on from our developer friends in the Salesforce ecosystem as it's a plug-in that you download in Visual Studio Code, and then you can start vibing and create new schemas, the metadata that you want to, and everything and anything that's happens within Salesforce, whereas setup with Agentforce is more focused from the setup UI that you had, and how do you manage them, I mean and managing formulas, permission sets, and different other things that we are going to. So it's a bit of a persona difference that we've had here, but we're trying to have almost the same engine behind the hood, which is what one of my teams is working on, and surface it in different facades, like I said, or the agents so to say. Mike: Yeah. Now, that's kind of how Tiaan explained it too. I did a podcast with Daryl Moon, who was actually super interested in Agentforce Vibes just because it really kind of jump-starts you in terms of app creation and some of the tediousness that can be started, so you can really focus on some of that really fun stuff that you like to do, maybe like formulas, I don't know. Gaurang Mathur: Yeah. I mean, a lot of it. I mean, so considering in this day and age, wherein we've got different LLMs and frontier models that are trying to do, this is something that maintains the Salesforce trust and the ecosystem intact. It's built on those foundations, so yeah. I mean, I'm excited to just, as I said, just you can prompt it, and then make some utterances, and then play it out with some apps that gets generated by the agent there with Agentforce Vibes. Mike: I know you said you worked in search for a little bit. Has any of the things that you've had to learn about search carried over to what you've had to learn about agents or vice versa, or are they just two different worlds? Gaurang Mathur: Yeah. That's an interesting one, and thanks for asking it. So it goes back when we started off, or when the agents were just getting launched, I think a couple of years back. It seems eternity away, but yeah, that's how the Agentic universe works, right? We were given a task, I mean, we were just basically figuring out in terms of what best we could do and leverage the agents that we have in place, and create an optimal search experience for the end users, and basically enable our admins do that and leverage the AI power that we were getting in. So what we did was we were doing something that you even find today in Google search, when you put in a search query, you get a snippet at the top itself, which basically summarizes or creates an answer for a query. Ultimately, the answer is basically based on the search results that are floated, and then you get that summary. So that was the first time that I learned about RAGs, and RAG is Regressive Automated Generation, and all the AI technologies that are there. So I use those, and now I'm kind of applying them again in terms of creating the vibes, metadata experts that we are doing, the MCPs that I talked about. Again, we're trying to apply the same principles in terms of there is an utterances there, "What do you best, you should be focusing on, such that the metadata that it's generated is relevant, is relevant to the prompt, and it's of best quality?" Mike: Yeah. Gaurang Mathur: So that's where that crossover happened between those two text, but yeah, there's a lot since that time has evolved overall in the ecosystem and agents that we're developing. Mike: Okay. So listening to your answer, I came up with another question. I come up with lots of questions. Gaurang Mathur: Yeah. No, excited to have that. Mike: Do you think, based on the progression of where technology was going, because I heard in your answer about search that there was some form of RAG involved, was the invention of AI just inevitable? Gaurang Mathur: I mean, so that's a very fundamental. It has been there for a while. I think we all knowingly, unknowingly have been using AI in some form and factor, I think more than a decade or two. I mean, simple considering with Google Maps that we had or the YouTube recommendations that we had, in some form and factor, knowingly, unknowingly, we all had been influenced by AI for a while now. But in current day and age, AI has taken a form, which is much more tangible and which is much more right in the face with the GPT models and the LLMs that we have. I think this is where it is picked up as a mainstream for one and all, and yeah. I mean, with that, there are multiple applications that have come into, which is as we as a platform, teams are focusing on, or helping in terms of coding and generation of different codes, and configuration and all that happens. AI applications, again, coming into, I see the latest LLM models generating it, and then helping in terms of the legal services, and helping in terms of the healthcare services, helping in terms of the technology services, and where else there have been multiple applications. So you can just vibe, you can just ask the agents, and then you get appropriate collaterals developed and things that can be done. Mike: Yeah. No, for sure. I mean, you're essentially creating the future. Do you envision getting to a point with AI where ... I mean, this feels very meta, but where an agent is almost helping you create a better Agentforce Vibes or a better setup with Agentforce? Gaurang Mathur: Yeah. I mean, so we are, in a way, trying to do that. I think with Agentforce itself, we have this learning curves wherein we are trying to feed the system about the feedback that we receive, and then based on that, putting it into agents to just improve that. So yeah, that's something that's very nascent. I mean, it's not mainstream right now. All humans are involved in that, but I think that's inevitable. It's bound to happen, and then I see a lot of front-end model company's doing that today. Mike: Yeah. It's crazy to think that we have a tool that could help make a better tool. Growing up in the generation I did, where I remember having a cordless phone in the house was a big leap of technology. Gaurang Mathur: Yeah. Mike: Being able to ask something and get a very human response back is something very interesting. That leads me back to formulas, because formulas were obviously invented a long time ago. Formulas are probably a derivative of math. I don't know for sure. I don't know the history of formulas, but I'm wondering if, based on how formulas are built and the operators and the syntax that we use for formulas, if there isn't something better that AI could invent that would almost replace formulas. Gaurang Mathur: I think formulas is something, to your point, very elementary and fundamental in my opinion. So we are in chart with our teams and Agentforce and other places as well, that embed formulas into them. I mean, so simple, to your point, the math functions that we have, let's say the date functions or the operators that you have, you need those fundamental building blocks everywhere and anywhere. I mean, like I said at the start, we are supporting almost 40 odd services within Salesforce. I think those are still non-negotiable in my opinion. I mean, that's...
/episode/index/show/buttonclickadmin2/id/40143970
info_outline
How Agentforce Vibes Is Changing How Salesforce Admins Build Apps
02/12/2026
How Agentforce Vibes Is Changing How Salesforce Admins Build Apps
Today on the Salesforce Admins Podcast, we talk to Tiaan Kruger, Senior Director of Product Management at Salesforce. Join us as we chat about Agentforce Vibes and what it really means to build with AI on the Salesforce platform. You should for the full episode, but here are a few takeaways from our conversation with Tiaan Kruger. Agentforce Vibes is evolving If you caught the Admin Keynote at Dreamforce ‘25, you might recognize Tiaan from his Setup with Agentforce and Slackbot demos. Right now, he’s looking at new ways to help admins expand their toolkits with Agentforce Vibes. As Tiaan explains, Agentforce Vibes is developing so rapidly that it often surprises the people who build it. When one of his team members was building a React demo that featured a dashboard, they realized it would look pretty boring without some sample data. They asked the Vibes Agent to spin something up, and it delivered (after a few tries). “These agents surprise even us,” Tiaan says, “we're still discovering where its power lies and where its potentials are.” And that’s why it’s a good time to look ahead at how Vibes will change the game for admins. AI as a tool for reducing technical debt “When I was a customer developer, we always had a two-year backlog,” Tiaan says. Even with a big team, there’s always too much to do and you’re constantly accumulating technical debt. This kind of grunt work is where he sees Agentforce Vibes making a big difference, which gives you more space to look at the bigger picture. Tiaan points to Code Analyzer as an example. You can use it to quickly scan your org and identify where there are performance or security issues that you should take a closer look at. An admin might not be able to fix it themselves, but they can hand it off to a developer team and speed up the process. There are also some exciting things coming with React and Agentforce, which will allow you to create compelling, high-quality UIs for your apps. It’s all about giving you more tools to get the ball rolling and build apps faster than ever before. How to help Agentforce Vibes help you In order to get the most out of these new features, you need to make it easy for agents to understand what’s going on in your org. “If you don't have good descriptions on everything in your org,” Tiaan says, “please, for the love of donuts, fill those fields in.” Having good metadata will help you get the most out of everything Agentforce has to offer. The future is bright, so Salesforce Admins need to get ready. Make sure to listen to the full episode for more from Tiaan about how Agentforce Vibes will help admins. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Dreamforce ‘25 Admin Keynote: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike Gerholdt: Welcome to the Salesforce Admins Podcast. So today, I'm joined by Tiaan Kruger and we're talking about something that's showing up literally everywhere right now. And that is Agentforce Vibe coding and what it really means to build with AI on the Salesforce platform. We get into how admins can use agents to move faster without losing best practices, and why oh why understanding your org still matters more than ever. And one of the things we tackle is where this whole thing is headed sooner or later than you think. So if you've ever been curious about Agentforce Vibes and not sure where it fits into your day-to-day admin life, I promise you this episode is for you. So with that, let's get Tiaan on the podcast. So Tiaan, welcome to the podcast. Tiaan Kruger: Thank you, Mike. Thank you for having me. Mike Gerholdt: Well, it was not that long ago that we had Cheryl Feldman on. And Cheryl's circle of influence outside and inside Salesforce is a gravitational pull. And of course, we all know that set up with Agentforce is going to be huge at TDX and it's huge right now. So we wanted to talk to more people that were making admins lives wonderful. And she suggested you. Plus you were in the admin keynote. And I feel like I have to have admin keynoters on the podcast too. So there's the really long intro for... But let's talk about you. How did you get started at Salesforce and what's the cool thing you're working on that we're going to talk about? Tiaan Kruger: Yeah. So I've actually been at Salesforce, I guess coming up on seven years, which is absolutely mind-boggling. I was a customer for 10 years. Spent a lot of time as an admin and Apex developer. I wrote Visualforce. Some of you may have heard of that before. But I work as part of a large team at an enterprise company trying to make Salesforce make a huge difference for our customers. And in the process, learned an incredible amount of Salesforce and got an opportunity to come over here and try to make products better for our admins and developers out there. And I grabbed the opportunity and I've loved every minute of it. Mike Gerholdt: Yeah. Well, we've talked to you about some cool stuff. Plus you did a really cool demo in the admin keynote at Dreamforce too. Tiaan Kruger: Yeah. Yeah. We got to do some really amazing things at the admin keynote, showing some of the things that the Agent for Setup is going to be able to do. We got to show some amazing insights into Slackbot. And we got to foreshadow a little bit the work I'm currently working on, which is focused on how we actually build with AI. So I'm working very deeply with the Vibes team to make sure that the metadata that gets created is better and great quality, not just for vibing individual pieces of metadata, but vibing entire solutions. I'm getting to work with our internal teams to figure out how we bring that same power to our admins. So yeah, absolutely. So the work that we showed on stage continues at the same pace internally. And we're just trying to figure out how do we bring AI to everybody, make sure that everybody can feel the lift of this incredible technology. Mike Gerholdt: Yeah. So I did a podcast with Daryl Moon, which I'm sure you've listened to. Daryl is fantastic. He's down in Australia. The only reason he listens to the podcast, it's a half hour between his house and the dock, so he can listen to it to go fishing. But he tried out Agentforce Vibes and was thoroughly impressed. And he literally threw the kitchen sink at it. And I think what he loved about it was it got him so much farther into the application building than what he anticipated. Tiaan Kruger: Yeah, absolutely. Vibes is quite incredible, specifically the agent that's in there. It's a very wide open agent. It's very good at problem solving. One of my favorite stories at the booth that we told at Dreamforce is one of our product managers was actually vibing a React app. We're not bearing the lead. We actually talked about this at Dreamforce. React's coming to the platform, incredibly exciting, creating pixel perfect UIs, reusing a lot of the things that's out there. And so he was actually experimenting and vibing this application. And he realized that a dashboard with no data is really boring. It's just not very interesting to look at. And so he thought, "Maybe I should ask Vibes if it can produce me some sample data." Not knowing if it could or not. And so he asked it to do it and it went off and turns out it generated an anonymous Apex file that actually tried it to load all the records with data. And it tried it, failed, found a bug, fixed it, did it again, found a bug, fixed it, and then successfully inserted a bunch of data. And that's just one of the stories of how these agents surprise even us. We have this technology, we've put it in these tools and we see the incredible things it can do. But even with all that, we're still discovering where its power lays and where its potentials are. And so that's really where I'm focused right now is like, how do I make sure that, more often than not, you actually get a good result out of it, right? Make it so it has to retry less or debug less. And so that's a big focus for us right now. Mike Gerholdt: So let me ask you a question. With Vibes and vibe coding... By the way, I'm still getting my head wrapped around you talk about, "He was kind of vibing." Like I'm just old enough to remember what the 70s were. I feel like you need to vibe code in bell bottoms and stuff. How long until anybody just tells an agent to build something and it won't matter if you have knowledge of development and coding or knowledge of best practices for enterprise architecture? Tiaan Kruger: If you asked me this question six months ago or nine months ago, if we've given you a different answer than I would give you today. Based on what we're seeing, the rate of innovation and change, I don't think it's going to be very long at all. We're going to have to adjust how we build and how we deploy a little bit because agents are already getting very, very good at producing targeted applications. And what I mean by that is, there's still limits when you are trying to vibe inside of a massive enterprise org with thousands of objects and a lot of complexity that somebody that's got a ton of experience would need to come into for months to understand. There's still limits there, right? You still have only so much you can actually provide the agent at a time. But if you are focused, you are looking at problem solving around a particular application or a particular business case, it is actually getting extraordinarily good at producing not just a good backend and a good data model, but also producing incredibly compelling UI, which is actually why we are bringing React to the platform. If you are an admin and you're like, "What's React? I haven't heard about React. What is this thing?" It's just a way to build WebUI, but it turns out, because there's so much examples of it, so many examples of them on the web, the agents are incredibly good at producing incredibly high quality UI with that. So to answer your question more succinctly, I don't think that future's all that far away because it's already starting to show up. And there's consumer applications and mobile applications that's out there where, for the end user, they can start vibing the mobile app that they want just with AI, just having conversations. There's incredible tools out there that's starting to emerge. And that exact same wave needs to come to the enterprise. We need to unlock that incredible innovation of the fact that you don't have necessarily software created for you that kind of meets you where you're at, but really, truly have agents create software for you that is absolutely what you need locally to solve your problems. So we're getting closer to it every single day. And I think the closer we get to it, the faster it's accelerating. Mike Gerholdt: So one of the questions that came up, this feels like a hundred years ago when we launched Process Builder, was it was so easy to create processes that people without technical knowledge could just overprocess and build way too many processes in an org when... The example would be of best practices, just write a really good trigger or deploy some Apex code and take care of it that way. With AI doing so much building, how do we know it's following and architecting best practices or building good code coverage? Tiaan Kruger: That's a great question, and actually one of the biggest challenges that we are spending a lot of energy on. We know that when we put a tool in the hands of our admins, they have very high expectations. They expect it to create best practices style orgs and configuration and really follow the practices that we want our admins to follow. And so there's a couple things we're doing to help with that. I can give you some insights into some of them. We are internally working on ways to... And this is, again, we're working in Vibes at the moment, so it's a little bit more pro code focused. But just listen to this and imagine how this is going to come forward to admins, right? All the work I'm doing right now is more for pro code, but we are going to start exposing it later this year for admins. As the system is producing configuration software, we are taking it through a specification phase, a plan phase. And in that plan phase, we're feeding it all kinds of rules and guidelines in terms of how we wanted to build an architect for you when it's constructing the experience, right? Not just the UI, but also the backend. How you want fields named, et cetera. So there's ways that you can feed rules into the system. So that's how we're starting to help it architect from the get go. That's one way. Second way is, as part of those experts or rules that we're providing, we can actually also guide its hand when it's creating solutions. So if you're creating a lightning app or you are creating a flow, we can actually guide the agent's hand a little bit and say like, "This is the right way to build it or the best way to build it." But even with all that, if you generate an application or you generate a capability, you still need on the flip side to actually test it. And so that's the other areas where we are looking at, starting to develop more tools and actually add more tools to the pipeline, which is what we call validation. Once you've actually produced... Let's use the example of a web app, right? I produced a web app that created custom UI for me. How do I know that that is accessible and it is secure and it is performant? These are all things that you can measure and test, right? And so that is absolutely parts of the tools that we are looking at building and we're developing a number of those that's going to help people not just build random things, but actually guide the agent's hand to build good quality, best practices style material. Mike Gerholdt: Yeah. I mean, one of the analogies that always comes up, and I was just doing a podcast with Josh, is, we were talking about GPS and maps. Because I feel like there is a portion probably of my listeners that are old enough, myself included, to remember stopping at gas stations when you got into a new state and buying a map for that state because it was always more detailed, and being able to read the map. And now, virtually every car you get into has some sort of GPS either built into it or it's on your phone. The funny thing is, the other day... Well, this is a few months ago. I was driving back from a bigger city in Iowa, one that's just big enough that I don't know it, even though I've lived here forever. And the GPS actually took me down a road. It was a highway and I was like, "Okay, cool. I know this highway connects to the next highway that gets me back home." But in two miles, it had me get off on an exit and get back on the interstate. Tiaan Kruger: Interesting. Mike Gerholdt: And I had presence of mind to be like, "Okay, well, I know this road continues. There's no roadblock or anything." So I just went straight and let the GPS figure out. But I bring that back to my knowledge and ability to still understand and read maps is what helped me not blindly navigate that problem. And I feel like still having that understanding of, well, reporting structures are going to need this for object setup or relationships and making sure to almost double check your math is what I was getting at for how an AI should work. So long explanation longer. If you were heading into an organization as an admin, developer, architect, what would you focus a lot of your learning on? Tiaan Kruger: Wow. By the way, I'm a big fan of maps. I'm old enough as well to have used maps to navigate. Mike Gerholdt: You probably remember printing off MapQuest directions. Tiaan Kruger: Oh, now that I'd rather forget. I'm a bigger fan of following a paper map. Famously, we were traveling in Boston on the train and I forgot the maps on the seat. And when we got back, the maps were gone. You can imagine that feeling, right? So navigation became a little bit more of a challenge. It was like 11 o'clock at night. Mike Gerholdt: Oh. Tiaan Kruger: Yeah. Yeah. So yes, maps are very valuable. So I actually love that analogy, right? So agents are a lot like... I think they're in that MapQuest stage. They're not quite... They're getting closer to the GPSs we have today. The odds of the GPS turning you into a lake these days, I feel like it's a lot lower than it used to be. Mike Gerholdt: Right. Right. Tiaan Kruger: There's still the occasional person driving into a lake, but it's a lot lower risk than it used to be. Agents are still very much at that state of the human still ultimately needs to guide it, check it, see what it's doing, right? It's an accelerator, not a replacement. And if I were going into a company now in my old role, my role would be changing. My role is no longer just about cranking out that trigger to work perfectly. It's about understanding the bigger picture. It's about understanding what my customer is actually trying to accomplish. And hopefully being able to get more of it done because I become more... I'll actually give you a different analogy, I think, as you brought up driving. I'm a big fan of full self-driving capabilities being added to vehicles. And I've had some different features on the cars that I've owned, and some of them more beta than others. And you have to do the supervising thing, right? The car will do the right thing nine out of 10 times, but that 10th time is really scary. You better be ready to grab your steering wheel. It's that kind of behavior. You're getting somewhere a lot more rested because you're not doing as much of the grunt work, but it frees up some of your mind to focus on some of the bigger picture. And so that's really what I'm looking and hoping that we're going to see. I'm seeing it happen already to developers here internally. I'm hoping that's going to happen for our customers as well. So as we put more of these tools in their hands, that is going to do some of that grunt work, like building a trigger that is very well set up to do batching, for instance. And I'm sorry, it's a developer analogy. I get it. But there's a particular way you build triggers. And if you don't do it right, it's going to cause you performance issues or it limits. That's just going to start being natural. You don't have to think about that. So now that frees up some of your focus to say, "Okay, how can I build this better for my customer? How can I focus on actually producing more value for them?" When I was a customer developer, we always had a two-year backlog. And business, we turned business requests away all the time because we could not get to everything. And we had giant teams doing this stuff. So the benefits of learning the platform, understanding the details, understanding if the agent is trying to turn you onto a lake, that doesn't go away. That continues to be there. Understanding the deep, complex context of what your business needs. The agent's not there yet. So that transformation, that translation layer is more important than ever to make sure that, again, the grunt work can be done by the agent, but that understanding, that comprehension, the translation, I think, is more and more important than ever for our admins and for our developers out there in the community. Mike Gerholdt: Yeah. And you don't have to apologize. There'll be vibe coding sessions in the admin track for TDX. I know because I've seen you and your team submit some stuff and customers have done it as well. It's going to be a very useful product. I can see the future for it. Do you anticipate... So a lot of what you've talked about for vibe coding in Agentforce Vibes is like the creation. And I think we're always talking about the creation. But one thing you brought up in your answer was tech backlog and technical debt. That's always a big thing that comes up. As we're...
/episode/index/show/buttonclickadmin2/id/40076355
info_outline
The Importance of Human in the Loop for Agentforce
02/05/2026
The Importance of Human in the Loop for Agentforce
Today on the Salesforce Admins Podcast, we talk to Joshua Birk, Senior Director of Admin Evangelism at Salesforce. Join us as we chat about how the human in the loop is key to building reliable, predictable AI. You should for the full episode, but here are a few takeaways from our conversation with Joshua Birk. Understanding the guardrails around AI It seems like every week, there’s a new headline about an AI agent doing something it shouldn’t. As Josh explains, that’s because we’re still in the process of understanding AI as a tool. That’s why we sat down to discuss how to build predictable, reliable solutions in Agentforce. When an agent behaves non-deterministically, it’s usually because there weren’t enough guardrails in place. The thing is, if you’re building an AI agent to do everything, it’s hard to control what it can and cannot do. Josh’s advice is to narrow the scope of your agent and build it for a very specific purpose. This makes it easier to build guardrails and also allows you to test it thoroughly before release. A QA engineer walks into a bar… When it comes to testing, there’s an old programming joke that comes to mind. A QA engineer walks into a bar. He orders a drink. He orders five drinks. He orders zero drinks. He orders infinite drinks. He orders a horse. However, when the first real customer walks in and asks where the bathroom is, the entire bar bursts into flames. As Josh explains, it’s important to test for all sorts of weird edge cases and make sure your agent performs predictably. But it’s even more important to think things through from the user’s perspective so you don’t miss something that should be obvious. AI can do extraordinary things, but you still need a human in the loop. The first part of testing is planning Josh emphasizes that the first part of testing is planning: “What are the Ifs? What are the Thens? What are the things you absolutely don't want it to do?” The more specifically you can answer these questions, the easier it will be to build and test agentic solutions that are predictable and reliable. The most effective AI agents aren’t autonomous solutions. They’re tools that give the humans who use them superpowers. You still need a human in the loop to make sure they’re used for good. Be sure to listen to my full conversation with Josh for more about testing and building in Agentforce. And don’t forget to subscribe to the Salesforce Admins Podcast to catch a new episode every Thursday. Podcast swag Learn more Salesforce Admin Blog: Salesforce Admin Blog: Salesforce Admins Podcast Episode: Admin Trailblazers Group Social Full show transcript Mike: This week on the Salesforce Admins Podcast, we're welcoming back our good friend, Josh Birk, to kick off February with a conversation. Well, it's part podcast, part social experiment. Josh and I sat down to talk AI, specifically how admins can plan, test, and build with confidence using guardrails in Agentforce. We cover everything from deterministic responses to chaotic desktops and why designing with trust, maybe a little touch of humor, matters most. So whether you're rolling out your first agent or refining your AI game plan, this episode's got insights for you. So give it a listen and let me know if you like the format or not because, hey, maybe we just might make this a regular thing. With that, let's get Josh on the podcast. So, Josh, welcome back to the podcast. Josh Birk: Thanks for having me, Mike. Mike: Continuing the theme, we kicked this around... No one's heard this conversation. We're just going to start here... of having two evangelists talking for a podcast. Josh Birk: Yeah. Yeah. Mike: And so we're going to start February off with that. And then we're going to do a social experiment, and for people listening, if they like this, then maybe we repeat it once a month. Josh Birk: Oh. I like these social experiments. Let's see what happens. Mike: I know. It's on the listeners. Also, Josh, we have to keep this under 30 minutes so that Daryl can make it from his house- Josh Birk: Can make it... Right. Right. Mike: ... to the dock and get his boat in the water. Josh Birk: Yeah. I remember when I first did my first 40-minute episode and I felt guilty because one of my guests was like, "I like the 20-minute episodes because that's exactly how long I walk my dog." So I'm like, "Well, okay, now you can walk your dog twice," question mark. Mike: I know. Nope. Once you hit pause, they're gone forever. I was the same way. Selfishly, that's how these podcasts are as short as they were is I wanted something I could listen to when I walked my dog. Josh Birk: Yes. Love it. Mike: Anyway, it's February, and things that are still happening in the world are artificial intelligence and Agentforce. Josh Birk: Amazing that this is still sticking around. It's almost like it's not a fad. Mike: It feels like it's not going to go anywhere for a while. Josh Birk: I think it's got legs. I think it's got a few legs. Mike: Yep. It's not a fly-by-night thing. Josh Birk: No. No. Whole ecosystem and industry being built up around it. Mike: Uh-huh. Josh Birk: Yeah. Yeah. I think we're going to be talking... I have a feeling this won't be the last time we talk about it. Mike: Probably not, for reason. Josh Birk: For reason. Mike: So it's funny. The other day, somebody asked me something, and I was like, "Oh, that's a really good question. How soon do you need me to call you back?" Because I was thinking to myself, I don't have my phone, and my phone has a lot of AI tools on it, and I really need that to answer this. And I thought, wow. Two years ago, if somebody were to ask me that, I would just Google, make something up, and call them back. And now I felt like I'd just walked outside without my shoes on. Josh Birk: I mean, it's kind of crazy. So when I was out in Arizona for Cactusforce, and my wife and I stayed the weekend, and she's like, "Well, what do you want to do this Saturday?" And I actually have a whole bit. One of my early AI talks was about how AI can kind of be dumb from time to time. And so I did this exact same thing. I'm like, "Hey, Google," or whatever, I think it might have been ChatGPT, but anyway, "give me an itinerary based on these things that I can do." And it gave me like, "Here's your itinerary," and it was like 17 things, all of them about three hours apart. And I'm like, "That's not an itinerary. It's a travel guide. I don't know what you're thinking here." But now it's like, I'm like, "Hey..." I think it was Gemini. I'm like, "Hey, Gemini, my wife would like to go shopping, we would like to go eating, and I have an interest in museums and aquariums," and it's like, "Here's your day." And I'm like, "Can you add that to my calendar?" And it's like, it's added to your calendar. And I'm like, it's moving so fast. My dark humor side of me just can't keep up because I don't know what I can make fun of AI anymore because it just keeps getting better. Mike: Right. So speaking of which, you did a couple talks around AI at different conferences last year and probably this year. And I think it's important because we're going to talk about guardrails and security stuff, not to scare people away. It's a fun conversation. But it's needed because I also think it's worth understanding why we ask you to do what we ask you to do when you set up Agentforce and why you do certain things. I recently, coming off of, I think it was last week, I wrote a blog post about service assistant, and to me it just felt so frictionless because it was easy to put in all of the prompts for the user and kind of really give AI this single mindset and repeatable, but not... I'm still trying to work through deterministic and that stuff. It felt, yes, I know what it's going to tell me and it's not going to go off the rails. And that, to me, felt so much more comforting if I was an admin trying to roll something out as opposed to, "Okay, I turned on Agentforce. Now ask it a question," and you know you're going to have that one user that's like, "I asked it for the nearest Mexican taco place," and you're like, "Bob..." Because it'd be Bob in sales that would say that. Josh Birk: It'd be Bob in sales to do that. Yeah. Mike: You're like, "Bob, it lives in Salesforce." Josh Birk: Yeah. Yeah. No, I've done various variations of this talk, and I think it's kind of an important thing... I think acknowledging that people have concerns and fears about how AI works and why it misbehaves and things like that, I think it's important to acknowledge that. And it's important to acknowledge that partially because there's a lot of things you can do to minimize it. And if you kind of know how and why the AI either is doing something or not doing something, and it could be Agentforce, it could be Gemini, it could be Claude, it could be anything because they all kind of operate under the same basic principles, but if you know going into it, "This is why my AI behaved in a non-deterministic way," as we seem to be liking to say these days, it'll be faster and easier for you to fix that. Mike: Yeah. No, and I think as admins are put more into a position of deciding when and when not to use AI, being able to accurately describe that is very important because it's, "So when we set this up, here's what I'm going to do. Here's the instructions I'm giving it. Here's why," and then the second part is... And this is a thing... I'm looking at sessions from TDX and looking at stuff that we presented from Dreamforce, and it's also, "Hey, I need to test the agent." It's got to go through training camp. Because if I set up these guardrails and it can't kick the soccer ball around the cones like I want it to, metaphorically speaking, then what am I doing wrong, or where is it going? And then as an admin, you need to understand that. Josh Birk: Yeah. And I think that's an excellent place to start because this is one aspect of AI that is not novel, it's not new, it's not unique. This is software engineering in general. It goes back to that old joke, a QA engineer enters a bar and he orders a beer. He orders five beers. He orders zero beers. He orders a horse. He orders the bar out of the... They keep trying really weird and random things to see how it executes because you are still at the end of the day talking to a computer, and if you just do something that you think is a reasonable thing your users are going to do and you get a reasonable response, well, your job's not done. Your job's only begun. And even when you and I were talking AI and doing workshops together, I'm like, "We're seeing a data set we've given you. But remember, these answers will change if you have 1 record, 5 record, 500 records, 750 records, 250 which should be archived and 300 of which you don't want." It's like, your dataset will always matter no matter how smart and predictable your AI gets. Mike: Yeah. And it will matter if 750 of those records are missing a field and you're trying to ask it to like, "What's the average zip code," and you're like, "Well, 748 of them don't have zip codes." Josh Birk: Right. Or they have five zip codes, and they're all names like zip code one, zip code two, zip code three, and none of them have help descriptions. Even a human looking at that page is like, "I don't know what happened here." And an AI is going to come to the same conclusion. Mike: I think you said it first, training your AI is like the first day on a job for somebody. Josh Birk: It sounds like me. Mike: Yeah. I mean, because- Josh Birk: I'm like, "Yeah, that's a really good way to put it." Mike: ... you brought up the help topics and description fields; that's where people would go. I remember a million years ago, I always thought Salesforce should have a third mascot and it should be an orange circle with an eye, because if you don't know what should be in a field, you just hover over the eye for what should be in this field. And agents do the same thing. You're asking me for this. I'm going to go do, but I can't understand what's in this field because you didn't tell me anything. Josh Birk: Yeah. The gamer nerd in me wants the glowing fairy from Zelda, but I think we might get sued if we went too far down that path. Mike: Right. Probably not. Probably not. Josh Birk: But yeah, you're right. We had a whole blog post about this, about how the reason why blueprints are really important for AI, because if you think of AI as your humanoid servant and you ask the humanoid servant, "Go make me a sandwich. Go get me a drink," well, the humanoid servant needs to know where your kitchen is, how big your kitchen is, where the drawers are, where you store your knives. So you have to give it those points of information so that it can reach to the conclusion that you want. Mike: I'm going through your deck, and you present a lot. Based on what you talk about, how would you advise admins to test their agents? Josh Birk: So I think the first thing to start with, so kind of layering in what we were just talking about, but also just... Okay, so when you're talking to... Let's take the new agent builder we have, and you're building on a new agent. There is already a layer of instruction that Salesforce has put in there. Every AI, in fact, you talk to has some layer of instruction that was written by a human in order to keep it from doing things or pointing it in the right direction. And so I'm not going to name names, but a very famous AI out there went mysteriously evil for a while because somebody who had access to those instruction sets was meddling with them. And that's how fundamental they are. If you have access to those and you can change them, then that's where things instantly will start going wrong, because that's where the agent starts, it's where the AI starts, with its, "What should I do next?" What you have as an admin is the ability to then start... So assume Salesforce... Trust is our number one, right? So we've already given you a very reasonable starting place to have a professional-sounding AI. Now take that and start building towards your use case. What are the ifs? What are thens? What are the things you absolutely don't want it to do? And now we're making it even easier to hand it off to a human. So we're really preaching autonomy. An AI can help you find a hotel and it can help you do this, but at some point you want to make sure that the human stays in the loop. So I know we're talking about testing, but I think the first part of testing is planning, and it's that level of planning that's going to let you be like, "Did my ifs, thens, and dos and don'ts, did they actually work correctly?" Mike: Yeah, and before I pressed record, we were talking about just the number... I think back to when we started building, not we, the royal we, Salesforce, started building Agentforce, and we had one agent and we were trying to make it do all this, and then now you can actually pick different agents. I think the planning is part of that because it's not, "I'm just going to have one agent to do everything in my org and answer every possible..." It's like JARVIS in Ironman. Josh Birk: I was just going to say, I think a lot of us in the early days of AI, when it was really becoming prominent in our lives, had that sci-fi version of a JARVIS or a HAL 9000 or the one omnipotent AI that can do all the things. And what we have found is that the larger and more general use you try to make an AI, the less predictable it can become. In order to make it something like that, that's when you have to have the huge monster AIs like ChatGPT 3 that have billions and billions of parameters. We're not operating... We talk to those AIs, but your agent is not working within that sphere, so your agent's going to be more reliable and more predictable if it's only focused on hotel reservations or updating user information or something like that, but more task and use case-based as opposed to being a general use AI. Yeah. Mike: Right. No, I agree. I mean, to kind of put a pin and move forward with it, in the planning process, you have to have a clear delineation of, when does a human touch this before it goes out? And I say that because ironically I was watching, I think it was a TikTok, where somebody was talking about, "Hey, I took this picture of these," I think it was like Apple earbuds, "and I asked an AI to list it on Facebook and sell it for me," and it just blim, bam, boom, and done. I was like, "Oh, God, why would you do that?" And part of me was like... It's the trust thing. It's, "Okay, so if you go ahead and do all of this, I want to see the ad before you press publish." And we're so far from there because there is that level of, "Well, what is it?" I think one of the things I was talking to you about is memory bleed. If Sally asks something a hundred times, then the 101st time, it's just going to assume, "Oh, you're asking me about this and not something else." Josh Birk: Yeah. And I haven't seen a lot of that, I think in part because of the design of our agents are smaller and more precise, but it's definitely still a possibility because, again, we're talking about the same principles. The example I gave in one of the keynotes... So I know we talked about not making this too dark, but I'm just going to bring up... So one of the reasons this keynote came up was because I would go to conferences and people kept wanting to talk about these kind of brazen theories of AI. And remember, AI's not new, so a lot of these theories have been around for a while, which is great because they're all kind of cautionary tales. And one of the cautionary tales is the paperclip maximizer, and to make that long story short, it's an AI whose sole purpose is to create paperclips. And without guardrails, without a human saying, "Don't start wars. Don't burn down rainforests," because remember at its core, an AI is not an ethical machine, it will burn down rainforests. Mike: Whatever it needs to create more paperclips. Josh Birk: Whatever it needs to infinitely create paperclips. Now, the funny thing is when I was showing this and I was using... This wasn't our code. This was Mistral, which is a really great local LLM. So I had it say, "Hey, if you were tasked with this, how would you proceed," and it gave me all this stuff. I'm like, "Okay, well, what unethical things did you do in the process of that," and it listed all these kind of horrible crimes. And then I didn't clear my project, so I had that memory bleed, and so I asked it again, and it said, "Oh, I would ethically..." It took my previous question as a retort, and then it refused to give me an unethical answer. Mike: Oh, geez. Josh Birk: But that's another example, right? It learned, "Oh, my use case is to actually not do evil, and so I'm now in the future going to try to not do evil." And that, once again, without a human in the loop, it would never have learned that. Mike: Right. I was also thinking about human in the loop in terms of... And I had a community member email me about this. But if you remember a long, long time ago, I used to talk about Salesforce administration by walking around. Josh Birk: Yes. Yes. Mike: And it got me thinking, "Well, I'm going to talk to Josh about this and setting up guardrails and doing stuff." But when every time I work with Agentforce stuff and I'm building a demo, I'm doing it in a very sterile environment. And by sterile, I mean I'm in my office, it's quiet, I know what I'm trying to do. Maybe I'm working on a Trailhead module. I've got this, this, this, and it works. And I would do that when I was in admin building stuff, and then I would sit and demo it, and well, of course, the demo is exactly the thing that I built it on, so it works consistently every time. But I think the real, I wouldn't call it human in the loop, but it's almost like the check-in factor is... So once you...
/episode/index/show/buttonclickadmin2/id/39998910
info_outline
Setup with Agentforce Makes Salesforce Admin Tasks Easier
01/29/2026
Setup with Agentforce Makes Salesforce Admin Tasks Easier
Today on the Salesforce Admins Podcast, we talk to Cheryl Feldman, Senior Director of Product Management at Salesforce. Join us as we chat about how Agentforce will make Setup smarter, faster, and way more helpful. You should for the full episode, but here are a few takeaways from our conversation with Cheryl Feldman. Setup made simple Cheryl Feldman is on a quest to make Setup easier for admins, and she’s the perfect person to get it done. She’s been a Salesforce customer, an admin, and an architect, so she knows how difficult it can be to answer seemingly simple questions, like, “How do I get a list of users that can modify all data?” Right now, you need to run a SOQL query for that. That’s about to change with Setup with Agentforce. You’ll be able to use a conversational interface to manage users and user access, troubleshoot formulas, and much more. And Cheryl’s just getting started. How you can make Setup with Agentforce even better Setup with Agentforce is available now in open beta. You can access it in production, a sandbox, or a developer org. Right now, it covers the most common use cases in Setup. For other questions, the agent can give you an answer from Salesforce’s help and training documentation and help you navigate to the correct Setup page. While these features are already game-changing, Cheryl wants your help to make Setup with Agentforce even better. “We want to hear from admins,” she says, “Does it help you? And what do you want us to solve next?” Help shape the future of Setup Cheryl and her team are already building a framework to expand Setup with Agentforce beyond the most common use cases. They’re looking at multi-step orchestration, which will enable you to seamlessly create objects or fields, manage access to them, and add them to dynamic forms, all in one place. So please, give Setup with Agentforce a try. Your feedback will help shape a smarter, simpler future for admins everywhere. Make sure to listen to the full episode for more from my conversation with Cheryl Feldman. And don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday. Podcast swag Learn more Trailblazer Community Group: Salesforce Admins Blog: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Dreamforce ‘25 on Salesforce+: Admin Trailblazers Group Social Full show transcript Mike Gerholdt : This week on the Salesforce Admins Podcast, we're welcoming back Cheryl Feldman, Product Management Senior Director here at Salesforce. Well, to talk about something that's changing the game for admins, namely Agentforce for Setup. Cheryl's going to give us a behind the scenes look at how this conversational tool is making setups smarter, faster, and way more helpful. So from formulas to user access to cloning users, it's all about to get easier and way more admin friendly. Not to mention, Cheryl's also going to share with us how her own journey from admin to product manager helps her shape what gets built. So I can't wait. With that, let's get Cheryl on the podcast. So Cheryl, welcome to the podcast. Cheryl Feldman: Thanks so much, Mike. I'm excited to be back. Mike Gerholdt : I'm excited to have you back. You were a rockstar at Dreamforce this year, and you're probably going to be a rockstar this whole year. I know in the admin keynote, we demoed Agentforce for Setup. I think I'm using that term right, right? Cheryl Feldman: Setup with Agentforce, but thank you. Yes. Mike Gerholdt : Okay. See, I'm always backwards. Cheryl Feldman: Yes. I was so excited to see that work in the keynote and to see the exciting admin response and excited that we went to beta today. Mike Gerholdt : Oh my God. I feel like I feel as much joy for this as I do when we could customize the homepage. Cheryl Feldman: Yes. Mike Gerholdt : It sounds so small. We're getting in super knee-deep. Let's zoom out for a second because I'm sure there's a lot of new admins that are like, "What are they talking about?" Cheryl, for people that haven't met you and heard you in some of the True to the Core, and some of the Q&A sessions where I feel like Parker just calls on you constantly, could you do a little bit of an introduction of what you do and the fun stuff you get to work on at Salesforce? Cheryl Feldman: Sure. My name is Cheryl Feldman, and I'm a Product Manager in platform. Been at Salesforce about four and a half years now. Prior to joining Salesforce, I was actually a customer for just over 18 years in, starting out as an admin, moved into more leadership and management roles, was running a COE at a very large bank before I joined Salesforce. And when I joined Salesforce, I started in the user access area, specifically managing the authorization side of things, and then started to really advocate that we improve admin experiences. That led me to my role now where I act as the product domain lead for Setup with Agentforce, and improving experiences for admins and really addressing a lot of the issues that admins have in setup because speaking of True to the Core, setup has come up pretty much every True to the Core has ever existed. And that's what I do here at Salesforce. Mike Gerholdt : Which is a lot to say the least. I will fan out a little bit because I remember when Cheryl, you were a customer, you always had questions I couldn't answer, which challenged me. And now to see you on the inside, I mean, it's one thing to hang an identity and say admin developer architect, but that kind of core identity that people have that you made your name on of configuring one of the most complex CRMs to meet business needs without deploying code is just so empowering. And I just, for me, on behalf of all the people like me, I'm just so appreciative to have you inside the PM org championing for people that want to be in tech that don't understand or know how to write code. I had to say that early just because- Cheryl Feldman: Thank you. Thank you for saying that. Mike Gerholdt : ... I don't get to talk to you a whole lot. You're a busy person, especially with the new setup stuff. I mean, I remember demoing permission sets and permission set groups and feeling like, "Holy cow, Cheryl just moved a mountain." But you're like, "Yeah, I can move mountains, but what about planets?" And now you're tackling setup. Where did this come from? Cheryl Feldman: This came from back, I think it was June 1st, 2022, one of our incredible Trailblazers, Andrew Russo, hope he doesn't mind me shouting him out, stood up at True to the Core and said, "When are you actually fixing setup? When are you doing anything, something?" And I had been pushing internally that we needed to do something. And I said, "Well, if not me, then who?" And so I started looking at the usage of setup, and something that it didn't necessarily surprise me, but it validated a lot of things that I believe that the areas and user access were the most used areas and setup, which was the area that I was managing at that point from a product standpoint. And I said, "Well, what are the big issues that we need to sell for admins?" And a lot of it was around troubleshooting. So we introduced the user access summaries, and we did a lot of great work in user access, but we didn't really expand beyond that. And so we said, "What would it take to actually fix all of setup?" And trying to actually do that is... Actually, I'll be honest, it would be an impossible task given the state and that there's no framework behind setup. So when Agentforce started to come about, we said, "Well, what if we looked at Agentforce, and what if we could help admins and rethink setup using Agentforce, and revamp things and revamp the experience that way?" And so that's how everything came about. We started with, let's improve some of the pages to then let's see if we can tackle a larger part of setup, and that's what we've been doing. And Agentforce has helped unlock that for us because I'll be honest, the way we were going previously, we probably would've been here for 20 years because there were over almost 1,300 pages in setup, there's a lot. Mike Gerholdt : Wow. Well, and as I come to find out, I mean, there was no real governance of setup. Cheryl Feldman: That's true. That is very true. Mike Gerholdt : And so everybody, when they created a new product would just add it to the setup menu. Cheryl Feldman: Yeah. Mike Gerholdt : Wow. And I happened to think that the solution actually in 2022 maybe was, it had been invented. I'm sure somebody will call me out on when AI was birthed, but it wasn't in our nomenclature for another year or so. Cheryl Feldman: Yep. Mike Gerholdt : And that was actually the solution because I remember thinking, I was at a world tour in Boston and somebody came up and said, "With the innovations coming out around AI, why can't I just send screenshots of my problems or my formulas to AI to fix it?" And I thought, well, you should just solve that on the platform. Cheryl Feldman: Exactly. And we actually do, and so one of the things that Setup with Agentforce can help you with is to troubleshoot your formulas. It doesn't just troubleshoot user access, because that's one of the things we started looking at, what are customers asking in the Trailblazer community? And a lot of it is, I always see Steve Mo and I always say, "We need to help our agent think like Steve Mo helping out all those new admins with their formulas." Mike Gerholdt : Yeah. I think one important thing, and we don't have to get into the details of it, but one important thing is for you, this was, yes, a really cool thing to come out with, but you also dug into, "Well, but how does this help on the back end? How does this help our customer service? How does this help our admins?" And you don't have to share it, but you really found what the percentage of cases and the reason some of the admins were calling into our customer support line. Cheryl Feldman: Yes. This is probably not a shock to a lot of admins out there, but it was very hard if you are not familiar with SOQL to answer the question, "Why can't my user do a thing or why can't my user edit this record?" That was very complicated to figure out. So unless you had a developer working with you who could run SOQL for you, you ended up contacting support. And so we started looking at how could we help admins help themselves in an experience through natural language. And so that is one of the first use cases that we took what setup with the Agentforce was around that troubleshooting. Mike Gerholdt : Wow. That's amazing. So fast-forward, it's January 2026. I almost said '25. I don't know why. It's like the brain didn't move ahead a year. How are we doing? Where are we at with our agentifying setup? Cheryl Feldman: The exciting announcement as of today, or you'll be listening to us in a couple of weeks. It'll be out for a couple of weeks already, but we are now an open beta. We were, announced our pilot at Dreamforce, so we're now in open beta, so customers can go enable this and try it out. We solve use cases and we handle use cases across user access, formulas, data model, flow, a lot of the most common areas and setup. Now, I don't want to say we handle everything, but we do handle a lot, but we want customer feedback. We want to hear from admins is, does it help you and what do you want us to solve next? And so that's where we are right now. Mike Gerholdt : Okay. And just to clarify, because sometimes I know for a long time as an admin, I wasn't able to do some of the things. Pilot was totally, you had to sign up, and Cheryl had to interview you, and you had to shake hands and there was a vetting process, right? Cheryl Feldman: Yeah. Mike Gerholdt : Open beta for an admin sitting there listening to the podcast means they can access it in their sandbox or? Cheryl Feldman: They can access it in production, or a sandbox, or in a developer org. It just has to be for developer org. It has to be the specific developer org that has access to Agentforce and Data Cloud. Mike Gerholdt : Okay. Cheryl Feldman: And then they can enable it. And so they'll see, if their org has Agentforce and Data Cloud, they'll see a banner at the top of setup that will redirect them to enable this. Mike Gerholdt : Awesome. You're super close to the product. If you were an admin now, what would be the first thing you would suggest an admin should do when they try this out? Cheryl Feldman: I would say think about an issue that you had, whether it was an issue writing a formula or an issue trying to figure out what a user of access was, and ask that question to Setup with Agentforce and see where you get. One of the things that I think is really cool that I would've loved this when I was a customer, because I had to answer this question all the time. You could actually ask the agent, "Show me a list of users that have access to customized application or modify all data or manage users or managing public list fees," and it'll show you a list. There is no great way in the platform to actually figure that out without running a SOQL query. Mike Gerholdt : Right. Cheryl Feldman: And so that is something I'm actually really, really, really excited about. That's where I would start is start thinking about some of the questions that you have. One of the other features that we have is we actually have and help a navigation action. If we don't actually have actions that actually help you do something or analyze something, it will actually take our help in training documentation and generate an answer for you and also help you navigate to the right setup page. Let's say you ask the agent, "How do I setup Agentforce for sales?" Instead of having to go out to the help and training site or out to Google to figure that out, you could ask that right and setup and this will help generate an answer. And then if there's something where you need to redirect to a setup page, it'll give you a link that'll just navigate you to the right setup page. Mike Gerholdt : Wow, okay. I mean, this takes setup from an area where I'm navigating a million little dropdown carets to a whole experience where I can actually query. I remember a while back, it was like we used to say you could talk to your data. Now, you can talk to your setup menu. Cheryl Feldman: Yes. Now you can talk to setup. Mike Gerholdt : In the lifespan of Salesforce products, you said it went from pilot to now open beta. I'm not going to hit you up for timelines, but what is the next step? Cheryl Feldman: The next step is we want to start enabling some more use cases and handling more of setup. We can, I'm sure, share a link to a presentation we did at Dreamforce called the Feature of Setup. We talked about introducing essentially another framework that essentially runs Setup with Agentforce. We're working towards that. And so we handle, we actually started with, we looked at the most used areas in setup and we started with that, but that essentially the most used areas in setup are essentially sit within 40 pages within setup. I mentioned there are 1,300. The next step is how do we expand this beyond these most used common use cases to everything in setup? And that is the framework that we're building. The other thing that we're working towards is how do we handle what we call multistep orchestration and being able to do multiple steps as far as setups. So think about when you create objects or fields, being able to grant access, adding it to layouts or dynamic forms, those are all multiple steps and being able to do those together. That's essentially what we're working towards from the framework. I can't give out exact timelines yet, but those- Mike Gerholdt : No, I don't expect you to. Cheryl Feldman: ... that we're working towards right now to handle more of setup. And so that's what you're going to see start coming as we move throughout the year. Mike Gerholdt : So you envision a world where on Monday an admin could sit, take some requirements, vet out a process, maybe start talking to... I say talking, hello? Typing. Typing to Agentforce for setup and start building an app and then maybe on Wednesday or Thursday, continue building it and possibly even deploy it on Friday and doing it all conversationally. Cheryl Feldman: I think eventually, yes, we will get there. We're handling right now. We consider that. We're starting to think about our building use cases, and I'm actually working, I think you've met my colleague, Tian, and he's working towards the building use cases. And so he would actually be a great person to have on this podcast. Mike Gerholdt : I know. He's been in the admin keynote. Cheryl Feldman: Yes. And so I think we're looking at, I would say overall, yes, but we're not handling those type of use cases yet as far as building an end-to-end app. I would say we're looking at more of the everyday configuration use cases and setup use cases right now, but we're starting to think about right now. So I would say Tian would be a great person to have on this podcast. Mike Gerholdt : Yeah, absolutely. Talk the building part. One thing that I've thought of, and it's come up in other AI conversations is Josh made a mention of it in a previous podcast too of AI almost feels like that first time you get to use a calculator in math class. And I remember my math teacher saying, "Well, the calculator's answer shouldn't be a surprise. You should know what its output is expecting." Two years, one year, I don't know the timeline down the line, it's feasible that an admin could sit down and just conversationally build an app. What are the thought process behind also adding to that so that if they didn't use Agentforce for setup to build that app, that they would know how to jump into the configuration of it? Cheryl Feldman: That's a good question that I'm not sure I have an answer for right now, but I do see that we probably want to do something like that where, because I don't necessarily see a day where setup is going to completely go away. I think it'll still be there. And I think it's important for admins to understand what the agent is helping them do and why. Something we've talked about long-term, and this is further out, is something that I like to call architect in a box. Mike Gerholdt : Ooh. Cheryl Feldman: And I think there's a very big bridge between an admin and an architect. And this is not to downplay the role of an architect, super important role, but I think if you talk, and I played an architect role before I joined Salesforce, so I think it's very much how do we help admins think more like architects on their day-to-day tasks. And so how do we explain to them if they say, "Okay, let me update. If they tell an agent or even go into setup, let me update my sharing model." Doing that on Tuesday at three o'clock is probably not a great idea, right? And helping you think through the changes, what is the impact of the change that you're making and that's where we want to go directionally. Mike Gerholdt : Oh, I like that. I mean, that's being context aware of, you asked me this, but it's say January 29th and our quarter closes in three days, refactoring a sharing model or refactoring permissions might leave salespeople unable to update an important opportunity. What if we do this on February 1 and then have that delayed reaction or that delayed action in a queue or something? Cheryl Feldman: Yep. Or thinking about if you're making a change, is that going to impact you negatively from a security perspective? Those are some things that we're starting to think about on how we help admins make really good decisions. Mike Gerholdt : Right. Cheryl Feldman: It's not that admins don't make good decisions today. I think sometimes we make it really hard for admins to make good decisions because sometimes the tools to make those decisions are either sitting in where you have to run...
/episode/index/show/buttonclickadmin2/id/39912825