Ep. 228 Building Trust in Agents: How Salesforce Powers Secure AI
Release Date: 04/01/2025
Federal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? In traditional commercial selling, a company seeks to understand its business problems and then presents a solution that would save time and money. Understanding federal requirements has a few more barriers than scheduling a meeting with the CIO. The federal government has security requirements and considerations few commercial companies can even consider. There are no effortless ways to understand system requirements for a company trying to break into the federal marketplace. This has been...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Most people are so overwhelmed with the dazzle of Artificial Intelligence that they dismiss the power of quantum computing. Reality is the optimized solution for solving some federal problems involving artificial intelligence; other issues need to consider quantum. Today, Murray Thom puts the ability of quantum computing in a better perspective. For example, when it comes to aerospace maintenance, there are so many variables that classical computing is challenged to provide an answer. We all know...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? It is a challenge for most technology companies to give a three-word summary of what they do, especially with the complexity implied with the phrase “entity resolution.” The CEO of Senzing, Jeff Jonas, gives a three-word summary of the complex issues they manage -- “bad guy hunting.” OK, what does this mean to federal tech leaders? Today, we sit down with Will Layton to learn how a topic like “entity resolution” can improve federal cybersecurity. During the interview, he gives an...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? For many, IDEMIA is a relatively unknown company that was recently awarded a 10-year Blanket Purchase Agreement from GSA. The focus is on next-generation identity proofing for login.gov. At first glance, you might say “IDEMIA” is an overnight success. Upon further examination, you will discover that IDEMIA has served federal agencies for over 60 years. During the interview, Donnie Scott gives listeners a complete rundown on the variations on identity, identity proofing, identity management, and identity...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Walking around the Salesforce World Tour DC can make you dizzy with use cases. Let’s step back and look at underlying principles. To boil down the massive information overflow, we sat down with Nasi Jazayeri from Salesforce to focus on improving efficiency by taking advantage of Salesforce agents. Automation. Federal employees will obviously be asked to do more with less. One way to accomplish this task is to structure a system where tedious decisions do not have to be made by humans....
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Federal leaders are walking a tightrope. They want to leverage the promise of AI; however, they are responsible for making federal data secure. Beyond that, these AI “experiments” should not negatively impact the larger systems and must have a detached view of practical applications. During today’s conversation, Paul Tatum gives his view on accomplishing this balance. He illustrates the idea of experimenting with AI through, of all things, avocados. For example, he acts as if he must document the process...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Most people know Cloudflare from federal or commercial experience. They have been around since 2009, and some will estimate that around 20% of all websites use Cloudflare for web security services. The listener's question is simple: can one apply this commercial success to improving federal network security? During today’s interview, Anish Patel from Cloudflare answered that question by directing his comments to Zero Trust, User experience, and automation. Zero Trust is a federal initiative that...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Once the transition to the cloud became dominant, the importance of identity was placed ahead of permission to access apps and data. When data centers were down the hall, one could have physical access to a room and sign-on permission. The hybrid cloud, private clouds, and an interest in “alternative clouds” make identity the keystone of modern computing. Companies like Okta, Ping, and SailPoint work with identity and access management but rely on services that can provide a federated identity service. ...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Many people deceive themselves when moving systems to the cloud, thinking the same precautions used for an on-premises system can be used in the cloud. Neil Carpenter from Orca Security dispels that notion right out of the box. He details that when a system is moved to the cloud, it operates under a shared responsibility model. While the Cloud Service Provider may be able to serve a solid infrastructure, that does not mean the applications and data are protected as well. Further, the popularity of...
info_outlineFederal Tech Podcast: for innovators, entrepreneurs, and CEOs who want to increase reach and improve brand awareness
Connect to John Gilroy on LinkedIn Want to listen to other episodes? Today, we sit down with Karthik Natarajan, Solutions Engineering Manager, U.S. Public Sector, for SNYK. SNYK has garnered a formidable reputation in the commercial sector by helping to identify and fix vulnerabilities in code, open-source dependencies, and container images. Karthik Natarajan acknowledges that no code can be 100% secure; however, one way to improve by a magnitude is to incorporate the “Shift Left” approach. This phrase has been around for twenty years but has recently gained momentum. ...
info_outlineConnect to John Gilroy on LinkedIn https://www.linkedin.com/in/john-gilroy/
Want to listen to other episodes? www.Federaltechpodcast.com
Federal leaders are walking a tightrope. They want to leverage the promise of AI; however, they are responsible for making federal data secure. Beyond that, these AI “experiments” should not negatively impact the larger systems and must have a detached view of practical applications.
During today’s conversation, Paul Tatum gives his view on accomplishing this balance.
He illustrates the idea of experimenting with AI through, of all things, avocados. For example, he acts as if he must document the process behind importing avocados. He shows how an AI agent can be used safely and provides practical information.
The key here is “safely.” People working on federal systems are jumping into AI agents without concern for compliance or security. They run into the phrase “unintended consequences” when they access data sloppily, which can lead to sensitive information leaks.
Rather than detailing potential abuse, Paul Tatum outlines the Salesforce approach. This allows experimentation with specific guidelines as well as for compliance and controls for autonomous agents.
This way, the data to be accessed will be cleaned and not subject to misinformation and duplication problems. Further, because you are acting in the functional equivalent of a “sandbox,” you can be assured that information assembled from AI experiments will be placed in areas where they are safe and secure.
Learn how to leverage AI, but learn in an environment where mistakes will not come back to haunt you.