CAIR 70: How To Reduce Your Business Cost With AI !!

In this episode, I have the opportunity to speak with Alex Hagerup who is solving the problem of using AI to take costs out of your business.

Grant
Hey, everybody, welcome to another episode of ClickAI Radio. So today I have someone that I have been admiring by looking at his background, here with me today to talk about some amazing aspects of his journey to solve business problems leveraging technology, specifically AI which is quite cool. Anyway, let me stop right there. And welcome Alex Hager group. Hello, Alex.

Alex
Hey, Grant. Thanks for having me. It's a pleasure.

Grant
Thank you. Thanks for being here today and for taking the time. I know you're getting ready to head off trans continental here pretty soon on a trip. So thanks for for jumping on this conversation here today.

Alex
Absolutely. It's exciting being able to travel again. So I going home to the Motherland for a few weeks is exciting. I haven't been there for more than a few days, for the last three years, actually. So I'm definitely excited.

Grant
Have have have the COVID situations and those numbers pretty good over there at this point.

Alex
Yeah, everything is fine. So Norway is completely open again. And it's all good. But But Norway was one of those countries that really looked down hard. And also since the US didn't allow non US residents to actually come back into the country. If you left. It was just a problematic situation to go to Europe in general. Wow.

Grant
Yeah. What a great opportunity to get home to family. Well, thanks for taking your time here with me today on this. So Vic AI. All right, who is Vic AI? What happened here? How to what are you? What are you solving? What what problem? Are you looking to address with Vic? Ai?

Alex
Yeah, absolutely. So I'll take a little bit of the background to set that up. So I, I've always been very interested in both accounting, finance and technology. So this is a company that lives in the intersection of those. My mom had her own accounting firm. So I grew up there, which probably influenced my interest for for accounting in general. And I built a couple of companies. But one that I spent three years with just prior to starting, Aki was a cloud ERP system, an accounting and accounting platform that was being used by about 30 40,000 companies back then it's about 80 90,000. Now, and during working there, you know, observe the sort of, let's say, the challenges of accounting and the manual, repetitiveness, the tediousness, all of that from, you know, every day you felt it?

Grant
Where? I mean, what is the excitement in that? Right? I mean, where's the excitement? Yeah. Did you credit that properly? 

Alex
Oh, gosh, yeah, exactly. So so we were we were just observing this. And then this was back in 2014 15. Just before we started with AI started having a new, you know, like a new summer or a new renaissance in a way and, and we were thinking like this has to be we can maybe like we can solve this in a better way than how the technology has solved it so far. And after some deliberation, we sort of thought that we could create AI algorithms that would be able to actually do accounting transaction processing better than humans.

Grant
Because that's a really key point I think you're making. So the way that we've been solving this problem up to this point has been, let me take the tasks that we do and just automate the tasks themselves. Right. In other words, let me take your actual transactional activities that you're working through, and just put some increased processing to that. That's how it's historically been done, right?

Alex
Yep. Correct. And that's been entirely driven by rules, right? So, you know, transaction, Uber, transportation, as, you know, rules based automation, right? And there are all sorts of problems with that. It's obviously better than doing everything manually, right? So we aren't we're progressing through stages there. If you go back before the spreadsheets, you know, everything was done entirely manually. So we are progressing here, but, but what we're building is not next face that comes off there, you know, what everyone is using today, all over the world. And, you know, AI will solve this in a more scalable and gracious way and more more effectively. So that was all right. Yeah.

Grant
Yeah. So real quick on that, can you articulate how is that different? Right? Because all of us come from this rules based way of thinking. So what is it that AI is going to do better? How will it do it differently than what we're used to?

Alex
Yeah, I mean, that's a that's a great question. It's kind of the essence as well. So when you when you look at how it happens today, it's not only rules are not only automation, it's automation, and a lot of human hours involved. So you can always ask yourself, why are all those human hours involved? If it's automation, right? So so it begs the question, right, so what happens in reality is that sort of rules and templates and RPA isn't sufficient, because there's so many edge cases, and there's so much variability in the world of accounting. So you know, rules only takes you so far. And then you have to staff up and have human cognitive reasoning step in and do the rest of it. So where AI comes in is that it can do both of those things. So it does the automation without rules, and it can do the reasoning that humans are there to do today. So I always say that AI is, is great at sort of mimicking that reasoning that humans are doing. So one of the areas that we're in is invoice processing. And when I give an invoice to any human, you know, it will always tell me, oh, that's the vendor, you know, that's the invoice number, lots of total costs. But that's not obvious to a computer. And if you're gonna write rules for every variability in the world, you can just end up writing too many rules. So it's just not a, it's not a great technology for it. And AI is way better. So it's just like in the early stages, so that's next sort of digital transformation, transformation journey that we're all. So as

Grant
So as you know, when you're working in the AI space, and you're saying, Oh, I'm going to apply AI to a particular problem, you end up building different models with different AI characteristics based on the nature of the problem, some more aggression based some more sort of classification based, as your people as your customers look to use your platform, do they get exposed to any of that? Are they even aware of what elements or aspects of AI are at play? Or? Or do they just jump in and start solving the problems they're used to, and then the right sort of AI model behind the scene is executing on their behalf?

Alex
Yeah, we've hidden all of that from our customers. So we do try to keep sort of Explainable AI in the way where our user interface is explaining why our AI predicted something. But we've kept all of the complexity, so sort of models and model training. And all of that is in the background, we decided to do like an end to end service where the customer, they don't really need to do anything technical. It's, it's, you know, just another SAS subscription that they're using, that they plug into one of their processes. And then we deal with all of the complexity of the models, both global models and AI models, specifically the same for each customer. So we keep all of that complexity hidden.

Grant
That's awesome. That way, I'm not touching any of that as the end user. So if I see the name Vic AI, I don't need to shy away and say, Oh, wait, I need to be an AI expert. It's more that this is the enabling technology. And it just turns out that you've made that simple for the people without needing to know that Wait a minute. So that brings a question on my use, you know, one of the challenges around AI is the whole notion around bias. And, and, and with the cognizance that's required for humans up to this point, and still largely today to do you know, counting processes. Therefore, that has the opportunity for some bias that comes in right terms of the way things get handled. How do you deal with that then in terms of applying AI so that that bias doesn't creep through?

Alex
Yeah, it's a great question. And it's a challenge for everything where we're data sets are involved in training, AI. So one of the ways that so one of The one of the good things with accounting data, if you start with that is that it's ultimately numbers and classifications. And you, you kind of want to have that, right? Because otherwise your your books of your business is wrong. So unless you want them to be wrong, you know, you have a very good incentive to keep this right. SO into SO, you know, I think we generally see kind of less less bias in accounting data, and then some other more like subjective data in a way. And then also we draw on data across 1000s and 1000s of customers. So we have customers in both Europe and the US and many other many other regions as well, but the little fewer and most of the concentration in Europe and US. And in all sorts of industries and all sorts of sizes, we have about 13,000 customers on the platform now. So when you start looking at such a wide data set, you also hopefully reduce some of some of that bias. And then you also have auditing processes that sort of sits at the end of their accounting. And hopefully also they'll you know, annually they'll catch corrections, and also fed in as well to make AI. So those are some of the tactics. Basically, it's all about keeping clean data, so that our predictions are accurate. That's really what we're trying to get to.

Grant
Yeah, boy, that's that's so critical, especially as you pointed out with the with the need to of course, be accurate for the business for sure. All right, so So let's say that you've got this data, you've cleaned it up, you've done the preparation, you're on the Vic AI platform, the question now in my mind becomes, therefore, what changes in the lives of the people that adopt this right and others, they may change? I'm supposing something about the way they do their daily work, or the way the CFO does certain things, or maybe even impact on regulations and audits. I mean, what's the impact of the organization when when this gets adopted? 

Alex
Yeah, I'm also in question. And I want to point out also that the beginning of getting getting live with Vicki AI isn't complex for a customer, because we built an automated system to ingest and clean their historical data, which then goes into our system automatically and train our AI algorithms so that when we go live, we have, you know, pretty good knowledge of what's going to be predicted. We know your system, we know your accounting, we know how you do things. So when you start pushing new transactions through our classification, accuracy is high. And there's no complexity for the customer. In that process, it basically happens automatically in the background. Once you are live with Vic AI, some things will will change. But it doesn't sort of change. Like overnight. One of the things with AI is it gets better and better over time. And one of the things we're driving towards is what we call full autonomy. And full autonomy is you know, the AI's version of automation. But it means that it has perceived this not to need human review. And that's when it's fully autonomous. So that's our sort of end goal with the autonomy of transactions is that, you know, the AI system is perfectly confident in the work that it has been doing. So it doesn't even ask a human to review it. So this this increases over time. So when you start with Aki, you, you have a you know, you have a better interface, you have a smoother operation, you have already probably 50% time reduction in the first month of using the system.

Grant
Let me stop you on that right there. 50%. So, that would be 50% of those that are doing sort of the daily operational activities, or is that of the CFO? Or who who is that, that that percent impacted?

Alex
Yeah, that's on the accounting team that is doing the processing. So if you're doing you know, if you have five people or if you have 15 people doing invoice processing, whether they are onshore or offshore, you know, just in the beginning, you can drastically reduce that, and then that percentage just increases over time because you're substituting the AI for for for human FTE. So basically, and and what we see is that everyone wants to do something else than just sit and do like data entry and accounting classification, right? You you, you can be more proactive, you can do more value added work than that. And we're sort of at that phase now where the AI can substitute that's partly augments and also fully run things autonomously. So when you put this in place, in the beginning, there is a little bit of effect right away but you know, you got to read the sign some of your some of your processes and some of your routines because you have a platform here now that is doing Most of the work for you, and you're just reviewing and you're reviewing the AI and training the AI to become more autonomous. So you've got a little on my mind shifts, and some sort of routines, you don't need to double and triple check, check in have four people involved in reviewing, you know, one thing because the AI can tell you how confident it is. And if it is very confident, maybe you can have one person review it in the beginning, and then eventually it will be fully autonomous. If it's less confident, then it will also tell you, and you can review it in more detail. So it's pretty, it's pretty, pretty fascinating.

Grant
It is dang fascinating. And I'm assuming that there are some that have run into this, and they've worried about their jobs, right? They're like, wait a minute, you're taking my tasks away from me? Do you have to help them overcome that fear and say, hey, you know, you're gonna move towards more value added activities within the organization? Have you run into that problem?

Alex
So it will we see is that it's, everyone is just squeezed on time, right? Everyone's trying to hit the deadline of the month, the clothes and all of that. So it there is no shortage of work to be done in the accounting piece. And, and just having, you know, having faster turnaround time having more accurate data, because AI is more accurate than humans, like it doesn't fat fingers thinks the same way. And when it's uncertain, it asked for a human verification, a human looks at it, and then you get more accurate than then all the way. So we see that, you know, there's no shortage of things to do. Everyone wants to progress their careers. And I don't really perceive that as an actual problem. But it is like a, you know, people think about that as a problem, but I don't think it is in reality.

Grant
So this has been several years and coming. When did you start this 2 to 3 years ago?

Alex
Yeah, early 2017, was 5 years.

Grant
Okay, that's, that's awesome. And you built this ground up, meaning all of AI development AI technology that your organization's created, right?

Alex
Yeah, exactly. I mean, it's been challenging. That's why it's taken taken quite a few years, as well as because you we started, you know, completely scratch, and we had to figure everything out. So one of the things that sets the KPI apart a little bit as part of our founding story, where we were able to start the company, we had access to a gigantic data sets of accounting transactions, and all corresponding documents to that. So that helped us just spend the first two years we just spent on data analysis, data science and machine learning development, because we had some thesis and theories that this could work, but we didn't know. So it just had to do that in the beginning, for for a couple of years. So when we saw that this actually has promised, like, we were predicting more and more accurately, and you know, we're gonna get to this inflection point where it's better to use PKI, that do not use PKI. And at that point, people will use it, and then, you know, continue growing more and more customers and more and more data and more and more corrections, and better and better predictions. So we realized we would get to that point. And then, you know, then we raised the seed round back in 2017. And, you know, started developing,

Grant
What a great journey. I love this story. So tell me about impact outcomes. So you talked about a large number of customers that are using the platform now, what is what's been the results that they've seen?

Alex
Yeah, so we see. So there's a couple of numbers that we that we statistically enroll from the from the customers, we see that customers have about 80% reduction in the overall time spent on the process. And that comes from two things. One is the percent of fully autonomous. So let's say you're 50%, fully autonomous, that means you spent zero time on 50% of your volume. And then that other part, we've drastically reduced the number of minutes in two seconds. For each transaction a human has to review because the AI has done all the upfront work, humans just reviewing it, rather than sort of processing it from scratch. So you're at seconds rather than minutes. This turns into sort of an 80% reduction in in time. You also have things that we do like prevent duplicate payments, and we have some fraud detection in the system. So you also have some of those benefits that can turn into multimillion dollar when you have a large enough cost base. And then we have audit trails in the system, which helps with figuring out you know, making sure that you know, all of the approvals are doing rights. If anyone has changed in amount or something. It's all logged into the system electronically. So you have some some compliance and auditing benefits from it as well. But right now the main, the main effective impact our customers have, and we typically sell to the mid market and enterprise. So these are larger organizations, they can have hundreds of 1000s, or millions even have sort of vendor bills per, per year, and a substantial amount of people and resources involved in dealing with that problem. So they see very significant ROI from it.

Grant
So to sell into that particular group, then I'm assuming you've got to have a decent amount of integrations into all of their incumbent systems, right, all of their ERP systems and CRM systems, etc. Right? What does that look like?

Alex
Yeah, that's, that's true. So we build out these connectors as we as we go, it does, for each connector, the first time we build it, it does take a little bit of time. And then once we have that connect, for instance, you know, to NetSuite or intact or Oracle, the, you know, the next customers that we bring on board, we can leverage the same connector. So it is it is some work initially to build out all of the ones that we need. And then you can grow and scale on top of that for the customer. This is no complexity like we're taking not on completely. So the only thing we need is to know what type of system they have. And we develop that connector as part of the offering.

Grant
So let's talk about looking to the future in terms of when I'm doing FPA, a financial planning and analysis and I'm, I'm looking at my numbers, and I want to leverage a view into the future. How does something like this help an organization with that?

Alex
Yeah, that's a, that's a beautiful question. And a big passion of mine, actually. So what we're doing now sets us up to solve that specific problem. So if you think of where we are, now we're in accounting. And what you're talking about is more finance, which is what we'll dive into the idea and the vision that we are getting to is that we want to develop an AI that's basically is a real time cost optimization engine that serves predictions and monitors that in real time, we will be able to help you project kind of what your cost base should be, and how you could reduce spend in various categories and with various vendors. That's that's the piece that we're trying to solve as part of that equation of the of the future. So that you always have multiple pieces of the accounting equation, you have the cost side, you have the revenue side, you have adjustments and close process. And we're trying to stay within one of them for now. And make sure we sold provide real true value in that swim lane before we move on from from there.

Grant
Yes, I think that part's fascinating, as well, I've seen and interacted with organizations that are trying to leverage shift P and A even into capacity and resource planning, and trying to figure out what that looks like, and especially with it what's going on in the world today, right, with certainly high inflation and certainly supply chain challenges. The need for this kind of capability, I think is dire. Right. I think having the ability to provide something like this sooner rather than later is really crucial.

Alex
Yeah, definitely. And I think the technology has the ability to hold much more context and see data across all of the things that buck forms, like ours can do in in, in a compliant and anonymized ways, you can see data patterns, you know, across different companies. So you can inform yourself, no more than just sitting looking at your own data sitting inside your own office and their own silo. So that I will, you know, in the future greatly help with, you know, capacity planning, or, you know, cost reduction initiatives and so on. So I think that's one of the powers of sort of, like the cloud combined with with AI and big data, if it's leveraged in the in the right way.

Grant
And so if I go back just one moment on something, so if we take an AI model, and and we take it through a data set, and we were able to get sort of two key views at it, one is sort of more hindsight, right? One sort of looking backwards, right? It's the AI is looking at it saying here were the drivers that that contributed to the certain behavior, right? So it's more analytic Right. But then there's the other style of using it, which is looking forward, which is more predictive. And oh, okay, here's, here's where based on things we've seen, here's what the high predictive sort of correlations are opportunities are things that we expect moving in the future. When you when you look at Vic AI, and everything that you're providing, do you focus more on one side versus the other? Or you're combining both of those views together?

Alex
Hmm...yeah, we so in the, in the course of the work in the platform we have now, so every thing we do is a prediction on new data. So that means a new so if you sort of break it down into like, there's a new signal or a new transaction, or a new document coming in, that we've never seen before. And, and the platform then predicts what this is, Donald to vary like line item level detail, how to classify it, and so on. So it does predict this based on the algorithmic design, and the historical data that it seen both for this customer and globally. So it sort of is our engines are purely predictive in terms of what they're trying to predict what's going on. And you can extend that to two other parts, like we just discussed on the cost side as well. You can predict how how you should, you know, maybe they'll just classify this cost, but like, how, what your cost base should consist of a new look at categories. And you can sort of predict this based on your own current spend your growth in the company and maybe other comparable companies, so you can start predicting a journey based on the same technologies.

Grant
Awesome. So quick question for you, as COVID started to hit, obviously, a couple of years ago, and large companies have their AI models, you know, out there deployed, we're running, suddenly that disruption created an impact to the way in which business was typically being done, right. Not all industries were hit. But certainly a lot of industries are hit. And therefore, business operations changed. And the way in which people conducted business altered, which meant then, that those models were making assumptions based on former operational models, right former ways in which people executed, and therefore invalidated some of those predictive characteristics or capabilities of those models, which meant then, as you know that they had to go do some rework, right had to reduce retraining, right and update the data sets and so forth. So to what degree does Vic AI continually learn, right, so as you pointed out, you get a new document you've never seen before. At some point, I'm assuming it pulls it into its corpus, right? And it continues to learn or relearn on that. Can you talk to that?

Alex
Yeah. And it's a key feature of the platform, which, which is actually pretty enjoyable when you see it working in reality. So we, so for every transaction that goes through our platform, or every invoice or vendor Bill divest process through our platform, that leaves that leaves us with some learnings. So, you know, it's either it's corrected, and we made a prediction and a human chose to change it. Right? That is something that is fed back into our data set, and also our learning database of sort of, why would that ever happen? What did we get wrong in this prediction? And so for every prediction that we do, we learn something because if we predict something, and it's correct, we also learned something. So you know, we have millions of transactions just flowing through the system every month. And for each of them, we learn something so the system then improves basically, for each transaction that goes through the system. The level of, you know, transaction processing we're at isn't really affected by kind of like global changes, like you mentioned, for for COVID. Like, the only thing that could happen is you suddenly have some some new transaction because you need to do something differently, or you may have fewer transaction, which then just you know, is additive to our system. So you know, we will get something maybe we'll have lower confidence on this because this is something new. So you know, you will have a human come in there and tell the system what to do with this. And then that will be additive to the to the data sets and, and all the algorithmic design, if that makes sense.

Grant
It makes total sense. Absolutely. Yeah. And I loved how you tied the two together which is, well, I'm still gonna have the transaction. Although the world the macro world around me is changing the transaction occurs while may change is the rate of them. or how, or maybe even maybe even the units or the number of units, etc, right things like that will certainly or could change because of disruption. Being able to predict or understand how the organization can respond to, to disruption, I think becomes more and more critical, as we seem to continue to have more disruptions in business. Right? That happens a lot.

Alex
Yeah, absolutely. And I mean, what happened with with COVID specifically was also a lot of disruption in how people worked. And, you know, they were used to working from home and, you know, we have some, some of our customers, they have millions of invoices a year, which means, like, 1000s of invoices a day and, you know, suddenly you have, your whole workforce is like, not as productive as they were when something like this happens. But you know, you still need or your your system to be updated, you still need to pay your vendors on time. And otherwise, you know, you'll incur fees, and it'll so just like having an AI system in there that doesn't sort of care about those things. It has the same throughput and outputs 24/7 all year around, makes you very resilient for for sort of things like that are like workforce changes.

Grant
So that's the right word resilient, you want that organizational resilience, and as the point I wanted to drive out, which is, even though there's these disruptions that take place, most of which are outside of our control, getting our companies into a position to handle and respond properly or well, to that, or to pivot is what organizational resilience is about. And my experience has been AI is one of those tools to help us do that. Yeah, that's, that's awesome. Okay, I want to ask you a very forward looking question. Right? You're ready? Yeah, sure. All right. So I'm not going to ask what are you going to do Norway? Or what I'm asking is tell me about blockchain? What does that mean to you in your world and things that you're doing?

Alex
Yeah, I mean, that's, we've been thinking about that. And, to be, to be honest, we haven't really seen yet the connection between specifically what we do and how to leverage a blockchain effectively, I like the technology of a blockchain is really awesome. And it can unlock a lot of things. But it doesn't mean that you need to use it for everything. One thing that I think could be an interesting exploration is when you start looking at accounting, ERP data, and whether the accounting systems could run and have their data in a blockchain that is verifiable, to prevent all sorts of, you know, fiddling with the data. That could be an interesting application for for blockchain in in our area. And there's clearly things around, you know, like money movement and value movements that blockchain can solve pretty pretty well. So just in the layer where we are, we haven't really found a great application that applies to us with the blockchain. Yes, but I think that technology is super fascinating. And I think that there are many areas where it is a superior technology to sort of like the trust based system that we have now. And it's an epic thing, when you can run that successfully in a resilient way.

Grant
When you think and that that makes a lot of sense. It's, it's got some, as you pointed out some interesting potential benefits to the industry. But I suppose at some point, it doesn't really matter where the transaction sits for you, right, in terms of the problem that you're solving, right? In other words, whether it's sitting on a blockchain or it's somewhere else and an Oracle ERP or whatever, right to it, at some point, that probably doesn't matter. Right?

Alex
Yeah, for us, the Store of the the sort of the store of the data ultimately is for us in the ERP system. And we are not an ERP system, we just layer on top of the ERP systems. So that is the ultimate sort of storage of the data. And that's where if, if, if at all the like blockchain technology could be leveraged to, to kind of store data in in a safe and transparent way. But there's also some new or some transparency issues with accounting data as well, that's going to be kept in mind.

Grant
Yeah, I would imagine that one of the greatest use cases to for what it is you provide is in the area of fraud to fraud detection, right the ability to regardless of where the data sitting, and regardless if it's on a blockchain or in an ERP system, you Your ai i would imagine would help to discover or uncover some of those potential opportunities. Is that accurate?

Alex
Yeah. And you have this this platform benefit as well is what if one customer sort of reports. So what happens a lot is just invoice fraud, right? You have, you have invoices that are not from the vendor that they appear to be from. And you have different variations of trying to change payment details, and you know, all of that. So when you see when you see data across 1000s of customers, you, you have a way of detecting those things, and at least flagging them for like, Hey, someone should have an extra look at this, because something may be up, if we're all good with it, that's fine. But you know, have an extra Look at this. And for people, you know, when we, as people, when we're doing something for six hours straight, you know, the last few hours, we're just like, you know, we're just clicking enter and trying to get, you know, the day over with, and, and things can easily pass by that shouldn't pass by. So this is this is problematic in two ways. Number one, your accounting data could end up being wrong, like you just have things that are misclassified. And then you can have things that are, you know, paid, or you're paid twice, or you pay the wrong thing, or you pay an extra zero, you have all of these things that are just humans are it's easy for humans to do those mistakes. And you just want to sort of get rid though those with a technology that can help you remove all of those errors. And we know how, you know, a billion dollar cost base? Just you know, half a percent is, you know, quite a lot.

Grant
A big number. Yeah, it is. Yeah, you'd hate to get false positives on a fraud situation, right, where you might be accusing someone that, that they've conducted fraud, when in fact it was was just a mistake. This is fascinating technology. So, Alex, where do people go to? I mean, where can you point them to, to learn more about you and your organization?

Alex
Yeah, I mean, the website is, is the safest place to go right now. So just go to Vic.ai. And we have, we have a good bit of marketing collateral and informative collateral blog posts, we try to educate the customers and the finance team. So finance teams are interestingly, you know, they're, they're not buying new things that often because you know, your ERP system, you have it for a very long time, which which you definitely should. So you know, you're not like a high. There's no high velocity necessarily on procuring things for the finance team. And, and we try to do a lot of educational material as well, because AI is new, but it has true effects or the finance team, and they'll find them blog, blog posts and collateral and can happily reach out to us and we'll help guide anyone through the process.

Grant
Oh, okay. That's, that's awesome. Alex, you've been so generous with your time here today, especially as you're getting ready to jet off into the sunrise just real quick. Any last comments that you want to share?

Alex
Um, I think this has been this has been a cool chats. It's been it's been really good. I love talking about, of course, AI and our company, but also the technologies in general. So I think this has been been super cool. And I think that it, you know, it's just wanted to say that we've we've spent a long time building the platform, so glad it's not complex for the customers right now, all of the AI performance, which is actually mind blowing, but you have no level of complexity. So, you know, if you if you want to get started with with AI for a function, this is a very specific function within invoice processing, where you can deploy AI, get the effects out of it with no complexities, and we'll deal with all of that for you.

Grant
That's actually the beauty of this, you know, to be able to get that down to a few points. Now splint clicks on a SaaS solution. They have all that taken care of, I've written my share of AI code, and oh, my goodness, the fact that you've handled that so seamlessly for the business people is huge. Nice job. Really great job on that. Yeah. Thanks. Great.

Alex
Thank you.

Grant
Thanks for taking the time today. Again, Alex. My gosh, this is awesome, everyone. Thanks for listening to another episode of ClickAI Radio. And until next time, go check out Vic.ai.

Thank you for joining Grant on ClickAI Radio. Don't forget to subscribe and leave feedback. And remember to download your free ebook, visit clickairadio.com Now.