AI ERP for Finance & Accounting teams to Replace NetSuite and Automate Close in 4 Weeks With Nicolas Kopp
In this episode of Future Finance, hosts Glenn Hopper and Paul Barnhurst sit down with Nicolas Kopp, the founder and CEO of Rillet, to discuss how AI is transforming finance operations. The episode dives deep into the practical applications of AI-native ERPs, with Nicolas explaining how his platform redefines general ledger management. The discussion explores the future of finance workflows, from the zero-day close to AI-driven automation in accounting tasks. It's a compelling conversation about integrating advanced technology into finance without needing deep technical expertise.
Nicolas Kopp is the founder and CEO of Rillet, the AI-native ERP designed to automate accounting and close books faster. Backed by Sequoia, Rillet empowers accountants by integrating AI seamlessly into financial workflows. Previously, Nicolas served as the US CEO of N26, a fintech bank valued at $9 billion, where he played a key role in leading its expansion into the US market. Prior to N26, he spent five years in investment banking at Morgan Stanley. Nicolas holds a BA from the University of St. Gallen in Switzerland and an MSc in Accounting from the London School of Economics.
In this episode, you will discover:
How AI-native ERP systems like Rillet are revolutionizing general ledger operations.
The process of automating complex accounting workflows with AI agents.
Why finance leaders need to embrace AI and the practical steps to do so.
The challenges and benefits of shifting from legacy systems to AI-driven platforms.
How CFOs can leverage AI today, even without a dedicated tech team.
Nicolas shared his journey from investment banking to leading Rillet, offering an inspiring look at how AI-native ERPs are transforming finance operations. His insights on automating workflows, achieving zero-day closes, and embracing AI-driven innovation provide essential guidance for finance leaders looking to stay ahead in the evolving landscape. This episode is a must-listen for professionals eager to drive change, innovate, and lead with purpose in the AI-powered future of finance.
Follow Nicolas:
LinkedIn - https://www.linkedin.com/in/nicolas-kopp/
Website - https://www.rillet.com/
Join hosts Glenn and Paul as they unravel the complexities of AI in finance:
Follow Glenn:
LinkedIn: https://www.linkedin.com/in/gbhopperiii
Follow Paul:
LinkedIn - https://www.linkedin.com/in/thefpandaguy
Follow QFlow.AI:
Website - https://bit.ly/4fYK9vY
Future Finance is sponsored by QFlow.ai, the strategic finance platform solving the toughest part of planning and analysis: B2B revenue. Align sales, marketing, and finance, speed up decision-making, and lock in accountability with QFlow.ai.
Stay tuned for a deeper understanding of how AI is shaping the future of finance and what it means for businesses and individuals alike.
In Today’s Episode:
[02:57] - What is an AI Native ERP?
[06:41] - AI Agents and Workflow
[09:15] - The Future of AI Agents with Autonomy
[11:47] - The Story Behind the General Ledger
[16:15] - The ERP Implementation Process
[23:43] - CFO Hesitations in Competitive Deals
[27:34] - Preparing Finance Leaders for AI Integration
[31:43] - Zero-Day Close as a Sport
[34:26] - Wrapping Up with Nicholas Kopp
Full Show Transcript:
[00:01:41] Host 1: Glenn Hopper: Welcome to Future Finance. I'm Glenn Hopper and as always, here with my partner in crime, Mr.Paul Barnhurst And our guest today is Nicolas Kopp. Nick is the founder and CEO of Rillet, an AI native ERP that's backed by Sequoia Real. It is reimagining finance operations with real-time visibility, automation, and a path to zero day clothes. Before founding it, Nick was the US CEO of neo Bank in 26, where he helped lead its U.S. expansion during its rise to a $9 billion valuation. He began his career in investment banking at Morgan Stanley and holds degrees from the University of Saint Gallen and the London School of Economics. Nicolas, welcome to the show.
[00:02:22] Guest: Nicolas Kopp: Awesome. Excited to be here. Thanks, guys, for having us. Yeah, we're.
[00:02:25] Host 2: Paul Barnhurst: Really excited to have you on. That's all. Glenn and I are always excited when we have our guests on, because it's always good to have someone smarter than us on the show. They help carry.
[00:02:34] Host 1: Glenn Hopper: These things really go off the rails when Paul and I just do one by ourselves and we just ramble on, we'll go on for hours and.
[00:02:41] Host 2: Paul Barnhurst: Release an episode. Yeah, we could go hours, and we're like, well, there you listen to two rambling finance guys like I. We hope you enjoyed it.
[00:02:49] Guest: Nicolas Kopp: Yeah. Love it, love it, love it. Let's make it. It's through. Yeah.
[00:02:52] Host 1: Glenn Hopper: And it's a good thing that you are so smart because you've taken on a monumental task. You know, you're not just going after one sector. You're building an AI native ERP. And I'm thinking about the times, and of course, everybody talking about AI. It's pretty exciting. But what does an AI native ERP mean? Like what's actually happening under the hood? And how is it different from, you know, say, a legacy system that just bolts on AI on their platform?
[00:03:17] Guest: Nicolas Kopp: Great question. And so yeah, for folks, this context where I give mid-market ERP is sort of leading the generational shift from the likes of the net suites of the world into the modern era in terms of the AI native ness. Specifically, it boils down to two key components where AI is really powerful today. So starting with the data ingestion, you have data flowing in from all these disparate systems, unstructured data sources in many cases, and research that needs to be done on transactions. And these are all areas where, frankly, AI and the technology that has been rising here is terrific as they can go through unstructured PDFs. They can research certain contexts on transactions. Um. It's really pretty mind blowing. We use mostly OpenAI's offer models. I guess we're taping this on June of 25. This may change over time, but I do a lot of that work. So I think that's really exciting. And for us, that part of AI is really natively embedded in the workflows of our customers. So the way data gets consumed comes in really like AI enabled. And then you have the middle piece of a general ledger ERP that's frankly not AI by design. If you want to run a depreciation schedule or you want to run a deferred revenue schedule to ground, you want an LLM nowhere near that, basically because you want deterministically always the same outcome every single time. So that's sort of I'd say the middle piece of an ERP that's by design, ideally not AI enabled. And then on the way out again you really have can use models again. And some of the advances there over the last 2 to 3 years to parse through information, compare information, summarize information. And that's where it gets really powerful again. But it's really a sort of a three step process where two of them AI is really powerful, the bookends, and then the middle piece. You have a good old deterministic code, frankly, that does the heavy lift for us.
[00:05:01] Host 1: Glenn Hopper : Yeah. And that's a significant advantage too. I just can't imagine thinking about how massive that like how many lines of code must net sweet Baby and going through and trying to figure out how to make it seamlessly work it. I mean, I could see them trying to just throw a bolt on gen AI on top of it, but to go through and really make it meaningful, it seems like it would be a fundamental re-engineering. And that's a it seems like it would be a tough spot to be in for the incumbents.
[00:05:28] Guest: Nicolas Kopp: Yeah. So without always trying to be correct and amicable with folks. So without maybe focusing on them too much but more like what we do. Very unique. It is very deeply ingrained in the AI in all the workflows. So traditional legacy systems are very much designed as proper general ledgers, databases, very much not in tune with the workflows that are happening end to end that these accounting teams do. And an accountant team will run maybe 2 or 300 disparate processes across multiple steps and workflows to ground that then land its way in a germ line one way or another at the end of their process. And so for us, the wave architect of the system is very much in line with these workflows. So again, with the way data gets ingested, the way data then gets calculated as well as output into all various forms, is very much in line with a smooth workflow. And then that allows us to also hold the data syntax itself, like so we control what information comes in, how it gets saved, how it gets structured. And that itself is a really powerful part of being a GL itself to basically control the data, but also the workflows in one. I'm happy to talk a little bit more about what we specifically do in AI, in terms of the actual agents that we have, but at least fundamentally or conceptually, this is how we're designed.
[00:06:40] Host 2: Paul Barnhurst: Yeah. I mean, what do you think about agents? I know this is a topic Glenn loves, so I love that you brought up Adrian's a little bit kind of how you're thinking about that workflow and agents and all that.
[00:06:53] Guest: Nicolas Kopp: Yeah. So for us. I agents are very much, um, synonyms of like, deeply ingrained, multi-step workflows that an accountant or finance professional would do on an ongoing basis. Um, and so, uh, examples of that are our, um, for example, cashback agents that we have are categorization coding agents that we have that 24 over seven. 365. We look at all the cash transactions coming in in the application. We'll scan for patterns. Um, context on these transactions and try to match and map these against either existing systems or create new entries to be approved for humans in the system. So that's a very interesting sort of like application of AI, where it's really helpful to have basically an authentic workflow to that work in multiple steps with a human would usually do, um, same for accruals, for example, where you have an agent and that can help you predict the accruals at the end of the month. It looks at all your transaction patterns, your bills in flight, and starts making suggestions for what accruals should be booked at the end of the month. And then once you, as an accounting manager or senior accountant, sign off on that, you hit go. And we actually booked a journal entry for you. So that's really a fully end to end automated process there which is once approved obviously. And so that is just like these are great examples of how agents can come to life very much deeply ingrained in the workflow. So it's not necessarily just a chatbot experience. It's very much like deeply ingrained in these like step by step workflows that accountants take at the end of the month.
[00:08:20] Host 1: Glenn Hopper : And I love that you still have at the end. There's still the human in the loop. We're not just turning over the reins and all that. And I, um, and you're spot on where I am. And, uh, in my day job, I focus a lot on building automations for accounting and finance teams and having that, you know, energetic workflows is the way to go. And you try to put as much as you can to the LM, but you can end up, you know, if they get stuck in the loop and there's a lot of error potential there. So you try to go as rule based as, as you can, you know, to prevent that from happening. But I'm wondering if we can hear Marc Benioff and all the marketing people out there who are selling their agents. How far do you think we are? And Paul, I'm going completely rogue from our designed questions here, but how far do you think we are?
[00:09:09] Host 2: Paul Barnhurst: I started the road journey. So your journey. I knew I'd let you go. And, Glen.
[00:09:15] Host 1: Glenn Hopper : How far from what you're saying do you think we are from truly a genetic behavior? And I'm not saying pull the human out of the loop and don't still have the approval, but sort of where the agent has agency and can think and act on its own and do things without being prompted.
[00:09:29] Host 2: Paul Barnhurst: Yeah.
[00:09:30] Guest: Nicolas Kopp: So for um, I'll give you again, you have all these software vendors running around pitching their solutions, and there's great marketing in there. Um, I'll give you a little bit of what we see hands on ground, because we use these models every day. We build these agents every day on sort of what works and what does so for us. Um, specifically, I'll start with what works. Well. What works well is simple. Distinct or discrete tasks. Breaking them down into, um, and multi-step workflows that you can then end to end run through. But the key, um, descriptor is that the important thing is like it is like scope down in terms of scope, where, for example, to that agent that we have in the platform, or we have an agent that helps run flux analyzes and goes through the entries and everything else. We pointed at very specific data tables because we own the GL, we own the data syntax and architecture. We can point it at the right data tables where to draw that information from. And same with other analytics use cases. We kind of break down individual junior workflows that would roll into maybe one specific person and scope it down, and then with that level can give it only a set amount of context, which makes the accuracy extremely high and also the speed manageable, as opposed to, um, some promises that I just don't see realistically work just yet is like, we give you access to everything and then the air magically figures out what to do. There's just too many loops to hang itself, and we all know that cleaning up a mess takes much longer than, frankly, maybe the automation time you save up prompt. Um, and so for us, we see we're maybe at 5% like where we can go end to end. And these like very broad use cases, the narrow use cases, we're like we're pretty far progressed and it's pretty exciting.
[00:11:16] Host 1: Glenn Hopper : That's awesome. I'm done Paul I'm.
[00:11:22] Host 2: Paul Barnhurst: It is exciting. So you know stepping back a little bit loves to get a little bit more of the backstory. So there you know how you came up with the name and how you decided to start it. You know most people don't wake up and think, I want to go tackle the general ledger industry. That's the business I want. I want to go up against Oracle and Sage and these well-capitalized, big established players. And, you know, most people don't think the exciting business is a general ledger, right? It's not the area, unfortunately. Most people don't realize it's sexy, but that's another story. So give us a little bit of the background, the name and kind of how it all came about. You decided this is what you wanted to do.
[00:12:01] Guest: Nicolas Kopp: I'll start taking you guys back a little bit to sort of the founding moments of the company. We can talk about the name shortly as well. Um, but I think there is, I mean, there's a huge opportunity in the market, I think let's start there. I think even most people would recognize that actually, like just on a white, you know, white paper, like a type of PowerPoint deck, like, oh yeah, that makes sense. Someone should build something like that. I think the hard part is really having the unique insight on how to get started. And I think we were pretty unique in that way that we just always like out of the gate. Our ambition was to build a full GL from the ground up, full stop. There was no cuteness. There were no shortcuts around this. Um, and at the time when we started a company at the age of 21, and then probably for the first two years of the company. That was a fairly controversial decision. Rippling was the only platform that you could kind of point to. Someone else has done a platform build here successfully. But outside of rippling, there were really not that many companies. So I think that's one thing we kind of got right from the outset, and it was very much personally informed by pain points that I had at my previous company in 26, where running some of these functions and we were looking at Workday Financial, NetSuite, sage intact, all these legacy systems there. And we have at the time, I think, five different ledgers, nothing consolidated in sort of.
[00:13:18] Guest: Nicolas Kopp: Any.
[00:13:19] Guest: Nicolas Kopp: Meaningful time frame, like everything was weeks and weeks. So, um, it was a real pain point that I saw back then again, combined with the opportunity, it became pretty clear that there is something there. And then I think, more importantly, at a more personal level, as a founder or also for your listeners, maybe an early employee joining a startup, you then still need to have the emotional guts to kind of pull the trigger on the idea. And so for me, the two things that I think kept us going and started is like number one. Um, it's intellectually extremely interesting. Like, how much cooler does it get? I know for most people it's very boring, but like, how much cooler does it get then? Having the ability to go build a GL from the ground up. No team has done this in like 20 years. You need to figure out everything yourself. So I think that was intellectually just extremely stimulating with my background of accounting and finance through and through. And then the other thing is you just want to build something that's meaningful. For me, it was really important that whatever I put my time to, which is starting a company is really hard, will have a meaningful impact over the next 10 to 20 years if we're successful. And so that was the other criteria in the early days where we just said, like, this is going to be really meaningful if this works out. So, um, and so it's kind of paying off and we're excited about that.
[00:14:32] Host 2: Paul Barnhurst: That's exciting. And the second part is really the name. How did that come about?
[00:14:38] Guest: Nicolas Kopp: So, real or related? Um, it's ironic that as a Swiss native and a first generation immigrant here in the US for the last 8 to 10 years using English terms from dictionaries. But like so it's really there really is a stream of water. Some people know it, some people may not be aware. I learned it over time as well. So really, really it is a stream of water. We pride ourselves with bringing financial information together from all these different disparate systems upstream into one consolidated source of truth. Um, and that was sort of the idea behind the name, the funny part there, the funny inside story there is we started with a completely different name. When we started out, the company was called now, which in my native language was a term for precision and accuracy. Um, and then as part of our fundraise, they're our investors. They were really excited about the idea and team, but they were very clear that we had to change the name if we wanted to be anywhere near successful here as we're scaling. So we pivoted over to really within the first 1 or 2 weeks of the company.
[00:15:35] Host 2: Paul Barnhurst: Thank you for sharing. I appreciate the backstory. I think I actually, I think you told that at the dinner I was at with you in December. I remember hearing that. So, Glenn, you wouldn't know this, but he came to Salt Lake and I had the opportunity. We had dinner with a bunch of finance leaders, and he helped lead it, and got the opportunity to chat with them. And so I remember hearing a little bit of that story.
[00:15:53] Guest: Nicolas Kopp: Yeah.
[00:15:54] Guest: Nicolas Kopp: And this is maybe actually a sort of a first time reveal on the podcast. Usually we tell the story behind, behind closed doors. Great.
[00:16:04] Host 2: Paul Barnhurst: We will definitely make sure that's, uh, in the trailer or a clip. Breaking news. Reel it and the back story. Yeah, that has not happened. I don't know how. First, I'd like to ask, what's the typical kind of implementation process like, you know, how does that work for customers? What should they think about that? Because I know, you know, generally ERPs are not known for being quick. And implementing can be a very lengthy process. And everybody laughs. I'll you know, obviously, I'm speaking to two people that have experienced that more than once. So maybe talk a little bit about that, how you think about implementation and then would love to just learn a little bit of what's on the roadmap for you guys.
[00:16:44] Guest: Nicolas Kopp: So implementation is obviously a big one. Um, nobody I think will be hard pressed to find a single person that enjoyed ERP implementation experience ever. And people were like a sign of pride, basically as a battle scar somewhere like, um, um, I've done five implementations in my life. So, um, in terms of the way we go about it. So first and foremost, it's 4 to 8 weeks. Depends on how complex the business is for easier businesses. We can even press it below the four week count for more complex business. It may be even slightly above, but the sort of the range we give is 4 to 8 weeks. And that's again today, June of 2025. We all have very, uh, cool paths and ambitions to bring this down into like a short amount of weeks only, um, and then, uh, the way that process works, and that is very unique in terms of when you compare it to the market, is we have an in-house team, all controllers, auditors, CPAs, that is enabled by technology, proprietary technology that we've built some of an AI, some of it not. I'm just optimistic in roles where we can, in a very quick process, get all the information out of your current systems. Have you then as the customer review your chart of accounts, department structure, everything you always wanted to clean up but never had the time for, and then basically run a full automated reconciliation of the back of that that gives you a sense of like old versus new system. Is the data correct? How does that look? And that process in itself, frankly, is a value proposition that is way and beyond what sort of legacy players do today.
[00:18:12] Host 2: Paul Barnhurst: Ever feel like you go to market teams and finance speak different languages? This misalignment is a breeding ground for failure in pairing the predictive power of forecasts and delaying decisions that drive efficient growth. It's not for lack of trying, but getting all the data in one place doesn't mean you've gotten everyone on the same page. Meet QFlow.ai, the strategic finance platform purpose-built to solve the toughest part of planning and analysis of B2B revenue. Q flow quickly integrates key data from your go-to-market stack and accounting platform, then handles all the data prep and normalization. Under the hood, it automatically assembles your go-to-market stats, makes segmented scenario planning a breeze, and closes the planning loop. Create air-tight alignment, improve decision latency, and ensure accountability across the team.
[00:19:20] Host 1: Glenn Hopper : I have to jump in here because I.
[00:19:23] Host 2: Paul Barnhurst: Now I just kept quiet. I'm like, Glenn is ready to go.
[00:19:27] Host 1: Glenn Hopper : Having been through a couple of painful ERP implementations and I won't throw out who they were with. But, um, I always say that the two quickest ways to get fired as a CFO. Number one. Fraud. That's easy. Number two, tell your CEO I think I want to do an ERP implementation. And you don't get fired that day. You get fired two and a half years later when the ERP implementation isn't done. So now, if there's ever anything that was going to make me go back to the office of the CFO, the idea of a 4 to 6 week implementation to make up for all those past injuries, I think that would get me back in if I could, if I could implement an ERP system that quickly. I think it brings me back. I don't.
[00:20:10] Host 2: Paul Barnhurst: Yeah. So get that in writing and there you go. But of course you'll pick the really complex company and then it'll be like, oh no, that one's a little longer.
[00:20:19] Guest: Nicolas Kopp: People always ask us out front. Like, how long does it take? I was just like, yeah, we scope these things and integrate an amount of rigor upfront basically. Sure.
[00:20:27] Host 1: Glenn Hopper : And that's I mean that's really amazing. And like you said, going against these big incumbents and it is a huge total addressable market. But it's also where your incumbents are embedded. And it's what's the old saying? Nobody ever got fired for choosing IBM or whatever.
[00:20:46] Host 2: Paul Barnhurst: Yeah. Ibm, Oracle SAP. Right. For using the incumbent basically.
[00:20:50] Host 1: Glenn Hopper : So the idea I think and you know, finance and accounting people are risk averse and all that. So it's got to be a hard sell to get them out. But I think when you talk about things like the short implementation and, and all that, but maybe and I don't know if you use this when you're talking to, uh, uh, potential clients, but Sequoia backed really early and that's a really strong signal in a very competitive space. And I guess so my question would be kind of what did Sequoia see, and what did these very risk averse CFOs see that made them willing to say, this is something we want to try or back in the case of Sequoia, yeah.
[00:21:32] Guest: Nicolas Kopp: I think that the decision or the thought process of an investor and customer is a little different in terms of the way they engage with us. So I'm going to start with maybe quickly around since you mentioned it. And then I'll go into how our customers think. So uh, specifically on uh, on the round itself, it is very much, frankly, sort of the leader in this in a space that is emerging here of like best product, best customers, most sophisticated customers, best team. Um, and so Sequoia has a very rigorous, frankly, decision making process. They were spending an insane amount, which is awesome, an insane amount of time with us leading up, even before the fundraiser, leading up to the fundraiser and during the fundraiser and turning every nook and cranny or like, what does it turn? Every stone? Uh, in English, like they went through everything in a great amount of detail. Spoke to a ton of our actual customers' prospects. Um, uh, references on the team. And went really deep on things. So I think for them, it's just a, um, a you need to get comfortable around the category itself. I think we could kind of show them why this is so exciting here. Um, in terms of generational shift, going from AC on prem to cloud, was NetSuite basically going from cloud to AI first? Now is it really it? so you have a huge market opportunity.
[00:22:49] Guest: Nicolas Kopp: And then as part of the market opportunity, you pick the best team basically as part of that. Um, and so that was sort of the decision making process of a funnel like that. And then for customers, more specifically, it is similar in the sense that they very much bet on a team as well. So we need to be realistic here that customers see that as much as they're buying an awesome product today, they also see the potential where this can go from here. Like it is, it's mind boggling in terms of innovation velocity. Again the AI agents that we have, the AI functionality like it's head and shoulders already above like sort of frankly what some of the legacy incumbents have done in four years. Um, and so I think that's sort of one thing. They very much take a bet on the team as well by buying the product. And then obviously, functionality wise, uh, we the functionality is pretty balanced and they really like what they're seeing. So that's the other, uh, element of that, combined with maybe also some other trust factors of who else are your customers and things like that.
[00:23:40] Host 2: Paul Barnhurst: Glenn, I'm going to sneak in here and ask one question. I'd love to know what's, you know, kind of the typical way you talk to CFOs and you're trying to close these deals. What are you finding? Kind of the biggest roadblock is it? Hey, you guys are new. Why should we trust you? Is it? You know, we can go with NetSuite. We're comfortable. We know. We know what they have. Or sage or whoever. What? You know, kind of what? That CFO. What's the kind of the biggest challenge or biggest thing you're hearing is you go through these competitive deals.
[00:24:11] Guest: Nicolas Kopp: What should I say before? Nobody's ever gotten fired over buying insert, IBM, Oracle NetSuite, all these names. And it's so true for our category specifically is it true for any software category. But I think for this one it is really exponential. Even more so.
[00:24:25] Host 2: Paul Barnhurst: True. So I would agree.
[00:24:27] Guest: Nicolas Kopp: Yeah. So it's really hard also to get started. So you can imagine that conversation with our first couple of customers where they like to tell us like, this is awesome, Nick. We like you. We like your team. Like the product is great. But like, how many other customers do you have? And then at that point, like if you have zero customers, it's really hard to like, um, get the first folks across the line. So, uh, the early days were really hard around that sort of dynamic. Um, by now we have a lot of reference customers in specific industries of our customers, so it gets gradually easier that part. But I'd say that's the number one objection point of course. But like, I know NetSuite is maybe the legacy product, but, um, I also know that you're forward thinking, but there are 40,000 of NetSuite, has 40,000 customers, there's 40,000 other businesses on NetSuite today. And so that's sort of like I think the dynamic that we're piercing through, where the argument that we're making in response to that is like, this is a really long term decision that you're making here. You can either not or we may not be perfect, we may not have the same penetration, but you see our trajectory and ramp over the next couple of years. Um, very clearly, versus something that is somewhat stagnant or has been stagnant for ten years. And that's sort of the counterargument to that. Nobody's ever gotten fired for buying. It's like, you may actually get fired for it in 2 or 3 years when there's literally yeah, when the gap is very stark all of a sudden.
[00:25:45] Host 1: Glenn Hopper : Yeah, that's such a great point because I deal with clients every day. That's why I'm sort of moving out of the SMB space because they just don't have one. The data that they need and or the resources and all that. So but at the mid-cap and even at the enterprise level, finance people are getting pressure from whether it's boards, investors, management, whatever to use AI, you know, sprinkle some AI on whatever you're doing and, and magically fix it. But for a lot of companies especially, you know, if we're just talking about the CFO's office, they're going to have to wait until the SaaS providers have AI built in. They're not going to go build their own tools and all that. You know, if you're at a large enough company and you've got a data science team and you've got machine learning engineers and you've got, you know, you've got developers that can do that. That's great. But for a lot of people, you know, finance isn't going to get that. And I think that seeing how quickly startups are coming around AI native as real, it is versus waiting on the incumbents to finally have some kind of AI included in their platform. That's going to be an easy way for them to check the box of, okay, we're using AI.
[00:26:55] Host 1: Glenn Hopper : We were getting pressured and now we are, but there's this gap even with that. And I always when I talk to clients, I explain to them that you need to understand. You don't have to become a developer, but you need to understand what's kind of what's happening under the hood. What's the difference between classical machine learning and generative AI and what an agent is and what an authentic workflow is. And just when you can, you know, when it's good to use AI and when it's good to use a more deterministic system. But I wonder, stepping back from your product just because you're so immersed in it. People ask me every day, you know, I'm supposed to be learning. I. I'm supposed to be experimenting. Do you have any specific things that you could say, even outside of your platform? Like what? How should finance leaders today be using and thinking about AI? And are there things that they can do to kind of prepare them for the future, where they understand how to interact with reality if they know where the data is coming from and whatever other tools they're using on top of it, and their whole finance tech stack.
[00:27:55] Guest: Nicolas Kopp: Definitely. There's one thing, actually that you said, Glenn, and if I may, I may actually push back on a little bit in terms of like, people may not have the resources. People may have the machine learning engineers or like the data scientists who go through this work. The thing that I would encourage your listeners on is like, don't wait for these teams. Don't use that as an excuse to not adopt AI here. Um, so I've, I've, I've had we host these dinners across the country and Paul was at one of them. It was awesome to have you. Um, thank you for joining. It was, uh, people are very scrappy and resourceful these days, and you can literally build an authentic workflow yourself with OpenAI just by tinkering with a bunch of prompts. And so we had this beautiful example just at a recent dinner where someone built a very specific workflow they had to pick up from a lock box, I don't know, like 50 to 100 checks every Monday or something. They built a little GPT thing where they could just feed that PDF, whatever. They would get into the GPT, and it would spit out a CSV file with like that, the amounts, the names and whatever other information was required from there, some reference numbers and it actually worked. It wasn't perfect, but it saved a bit of time. And so I think what I would just encourage folks is the way to get started is like, no excuse of waiting for someone else. Try and look at these tools yourself, and it may require a bit of time, but it's actually possible to build stuff that helps you save time. Yeah, it is a 2 or 3 hour afternoon that you would invest and that's this thing. Or like sit down on a Saturday and go do it. But it will save you time and you don't need to wait for others. So, um, for me, that's sort of maybe the biggest call to action for your listeners of like, try and start tinkering. It doesn't have to be perfect. Start tinkering.
[00:29:35] Host 1: Glenn Hopper : Oh, I was gonna say, I always tell people I don't know what the base level of ChatGPT plus or whatever, 20 or 25 bucks a month, go ahead and pay for it. It's that you spend more than that on coffee in a week. Go pay for that. It's going to pay for itself and then spend. I mean, the ROI. We're finance people, the ROI on something, even if it takes you. I worked with a group a couple of months ago, and some guy who had no developer background or anything, but got a written code for this, something that he was doing that took him an hour every day. He spent like 14 or 15 hours, uh, going back and forth for whatever reason. He was going back and forth between ChatGPT and Claude, but he was writing code to make a little worksheet. But he automated this whole process. And the ROI is simple. He spent 15 hours there. If it was an hour a day, that's three, three business weeks. And he's and he never has to do it again. Roi is right there. So I'm right there with you on experimentation.
[00:30:29] Guest: Nicolas Kopp: Yeah. And also to use, uh, finance terminology for your listeners. Also, if you discount the future return that you're getting because you're uncertain if it'll work out. Yeah, even with a discount, it is it is worth it if you start small and don't pick the.
[00:30:41] Host 2: Paul Barnhurst: Starter.
[00:30:41] Guest: Nicolas Kopp: Rate for sure. Yeah, exactly.
[00:30:44] Host 2: Paul Barnhurst: Love it. All right. So this is our kind of fun section that we surprise everybody with a little bit. We have 25 questions that have been developed by ChatGPT. Based on your profile, what I found on the internet, the questions we asked you to come up with some personalized, fun, unique questions. I haven't even looked at all. We never know what we're going to get. So here's how it works. Glenn and I both do it a little bit differently. I say pick a. You can do one of two things. You can pick a number between 1 and 25. And I'll read that question. Or I can use the random number generator to pick a number between 1 and 25. So this is as close as we get to humans in the loop on these questions.
[00:31:25] Guest: Nicolas Kopp: Let's go with three.
[00:31:26] Host 2: Paul Barnhurst: Three. All right. This one is under the section called naming and zero day close ideation. You already asked her questions. You can ask about the name of it. So I'm going to go to number four. This didn't happen to us before. This is a first. All right. You aim for zero day closes. If it were a sport, which one would it be? And what your championship is.
[00:31:55] Guest: Nicolas Kopp: Love it. And you speak to someone that is really out of sports in general. I love doing it myself, but I'm not generally following. So I'm going to use, um, formula one racing. All right. It is a zero day close. Be as fast as you possibly can. These drivers have an insane amount of precision when they go around the corners and, like, accelerate and do their little, uh, battles through things. Um, and, uh, whatever the grandest prize is in Formula One racing. I think there's actually only one, there, I think I guess we played for that.
[00:32:27] Host 1: Glenn Hopper : So, Paul, this is why you and I will never be a CEO or founder. We can't think this fast on our feet. We need scripts for everything. So well done, well done. Nicolas. So, um. Okay, so I do mine a little bit differently. I just because, uh, ChatGPT created the questions, I just get it to spit out and I say, give me one of these at random, and, uh, we'll see what it comes back with here.
[00:32:51] Host 2: Paul Barnhurst: You decide if you like it. Glenn has a little more human in the loop than I do. This is.
[00:32:57] Host 1: Glenn Hopper : This one is so weird. Oh, that's all right. That's all right. It says you once wrote a medium post about your first audit. If the audit itself could talk, what one comedic confession would it make? I don't think I was hallucinating here. I don't know Nicolas. If you've got an answer, great. If not we can go back to the drawing board. I don't know.
[00:33:22] Guest: Nicolas Kopp: I so there was a post I think that one of the first like content posts that we wrote was about going on the first audit. So that part is right. It's actually pretty impressive. And so what was the question that you asked about the post? In fact, if that post could talk.
[00:33:35] Host 1: Glenn Hopper : It says if the audit itself could talk, what one comedic confession would it make?
[00:33:42] Guest: Nicolas Kopp: Okay, I'll.
[00:33:43] Host 1: Glenn Hopper : I'll.
[00:33:44] Guest: Nicolas Kopp: I'll use a cheeky one. I'll maybe say like, please stop using that tweet.
[00:33:48] Host 1: Glenn Hopper : Yeah. That is perfect. That is perfect. I don't know, Paul. Every time we get these weird questions, I think maybe we need to throw the human back in the loop. But I sort of like just letting the I just come up with it.
[00:34:01] Host 2: Paul Barnhurst: Yeah, this is the one. This is the section where we get to show, okay. Sometimes maybe it makes sense to check, maybe not, maybe it could be fun. So we just go with it.
[00:34:11] Guest: Nicolas Kopp: Are you maybe checking the drum line generated by. We're at that level? Yeah, exactly.
[00:34:17] Host 2: Paul Barnhurst: So I honestly don't even read them. I just throw the 25 in the document and they go with the day we do it.
[00:34:23] Guest: Nicolas Kopp: So that's an awesome idea.
[00:34:26] Host 2: Paul Barnhurst: Yeah, yeah that's fine. Well, thank you so much, Nick Nicolas, we've had a great time chatting with you today. Love what you're building out. Really excited to see you continue to grow. Congratulations on the fundraiser. I know that was a ton of work. It's not easy to raise capital and to get Sequoia on the cap table is huge. So congratulations. You know thanks so much for joining us.
[00:34:47] Guest: Nicolas Kopp: Great to see you guys. Thank you.
[00:34:49] Host 1: Glenn Hopper : All right. Thanks, Nicolas.
[00:34:50] Host 2: Paul Barnhurst: Thanks for listening to the Future Finance Show. And thanks to our sponsor, QFlow.ai. If you enjoyed this episode, please leave a rating and review on your podcast platform of choice, and may your robot overlords be with you.