How Companies Are Losing Millions While Being Data Rich with Ron Nachum

Welcome to the Future Finance Show, where we talk about treasury management on bars.

I'm flying blind. Every single time I wanna make a decision, I ask my analyst for data. It takes them a week. I made the decision 3 days ago on where to spend the money. Companies have tens of millions of dollars locked away in their data warehouses. It's just about, can you find the way to get it out of there? The world is moving so, so fast right now. Like, the opportunity cost of not going and building something is so high.

It made it a no-brainer decision to just go all in and do it.

Do it. Meet Ron Nachum, co-founder and CEO of Sapien, the AI intelligence layer helping companies from mid-markets to the Fortune 500 make smarter financial and operational decisions. A Harvard CS and statistics student turned AI researcher, Ron has also led award-winning Alzheimer's research and headed Project Kylys as president and CEO, a student team building the world's first amateur liquid-fueled rocket to space. Multibillion-dollar companies get onboarded in 2 days. The first thing they get when they open the platform is not an empty chat box. It's, we already found 3 SKUs that are gross margin negative. We're not just building a chatbot, we're building this like system of action.

What's your dream for how humans interact with Sapien and how they're building this and using it?

Future Finance is brought to you by Qflow.ai, the strategic finance platform Solving the toughest part of planning and analysis. B2B revenue. Align sales, marketing, and finance seamlessly. Speed up decision-making and lock in accountability with Qflow.ai.

Welcome to Future Finance. I am Glenn Hopper, co-host along with my esteemed colleague, Mr. Paul Barnhurst. Paul, how you doing?

Doing good. Always excited to be here. How you doing, Glenn?

I'm good, except I mentioned before we came on, I'm wearing a sports coat and a t-shirt today. I feel like Crockett from Miami Vice. It's a black t-shirt, so I, for those who are just listening, it's a black t-shirt. So maybe it's like Crockett headed to a funeral or something. I don't know.

We're live so we could poll the audience and see how many people even know who Crockett is or recognize him.

That's true. I'm dating myself here by referencing an '80s television show that—

Hate to break it to you, Glenn, but you and I are getting old. Yeah. I mean, we're both the other side of 50 now.

Our next guest is not getting old. He's young and vibrant and everything that we remembered from all those decades ago.

I wish I knew who Crockett was.

See, I told you. Welcome on the show, Ron. We're really excited to have you here with us.

Yeah, I appreciate you having me here. Excited to dig in.

Yeah, we are as well. So I'll give a little bit of background about Ron. So Ron Natchman spent years in applied AI research before founding Sapien. Sapient is an AI-native system deploying agents that understand siloed, messy data to run, optimize, and act on company finances end-to-end. It sits on top of your systems of record. So ERPs, data warehouses, CRM, spreadsheets, et cetera. Ron studied, if I remember right, see, was it Harvard or Wharton? Harvard.

Harvard.

Studied at Harvard. He's worked on several AI research projects. And now he's scaling his company, which is powered by its company engine and tabular reasoning model. And we're really excited to get the opportunity to chat with you today. So again, welcome.

Yeah, appreciate you having me on. I mean, great. Couldn't have said it better myself on the introduction.

Perfect. All right.

You probably would've remembered that it wasn't Wharton though.

Yeah, you would have. What can I say? Those are all outta my league. I'm not like you, Glenn. Which one did you graduate from again?

I went to the Harvard on a budget. Is where I went.

See, here's where I graduated from for grad school, ASU, also known as party school. So there you go. That sums it up. Oh, my undergrad was BYU, which is non-party school. So go figure. So, you know, Ron, you and I have had the opportunity to chat a couple times already before the show. And, you know, something you mentioned I want to kind of dig into to start. When we chatted, you talked about, you know, so many companies are, right, data rich, but analysis poor. Why do you think that is? I mean, obviously some of it's messy data, but we just love your thoughts on what you're seeing and why that's such a struggle for so many companies.

Many years it's been true, but I think in 2026 we feel this gap more than ever. I think there's a lot of reasons why companies are data-rich and analysis-poor, but part of it is every company has had it kind of a little bit bashed into their head. They need data, right? They need big data. They need information. They need to be tracking as much as possible. And a lot of companies have actually been doing this very well. You look at even, you know, we work with lower middle market manufacturing clients that, you know, are generally not the most technically sophisticated. Even they are tracking a lot of this information. What ends up happening though is companies are, you know, data rich, but it's split between a bunch of places because, you know, they started tracking data in the '90s and then they acquired another company and now it's in a different system. They acquired another company and now there's different pieces all over the place. Not to mention the fact that the size of data that we then start talking about is, you know, tens of millions, hundreds of millions, billions of rows, the type of stuff that breaks Excel if you even try to think about opening it. And so what we end up seeing is companies have a lot of information, but they're not acting on it. And actually one of the first conversations I had when we started Sapien was with a former Fortune 500 CFO. And he told me, I'm flying blind. He was like, every single time I want to make a decision, I ask my analyst for data. It takes them a week. I made the decision 3 days ago on where to spend the money. And it's because large amounts of data, very, very siloed, way bigger than what humans were meant to work with. Like our human brains were not built to work with hundreds of millions of rows of data that doesn't agree with one another. This is where AI automation, all these things should be enabling us, right? Enabling people to do the work that matters. But I think we see it oftentimes, like— To answer succinctly, siloed information, way too much information. People aren't built to work with it. And what that means is usually we just skip it entirely. You know, we're not doing any of the analysis, or if we are, it takes 3 weeks to answer a question. So we're only picking a couple of questions to ask instead of asking every little thing that might move the business forward. We have to be choosy as a result, cuz it's all a resource allocation problem at the end of the day. So I think that's a lot of what we saw. It's actually a lot of what inspired us to really start Sapien, and it's a lot of what we keep seeing more and more. Everything is there. It's just how do you get the value out of it? We almost like to say companies have tens of millions of dollars locked away in their ERPs and their data warehouses. It's just about, can you find the way to almost like get it out of there?

Can you unlock it?

Yeah.

Right.

All right. So my first question is really just an, an observation. I'm gonna read you a list. Bill Gates, Mark Zuckerberg, Matt Damon, William Randolph Hearst, Robert Frost. You know where I'm going. And Buckminster Fuller. All famous Harvard folk who did not finish. And as I think about that list, I'm wondering, I mean, arguably that's probably rocket fuel for starting to becoming a billionaire industrialist, or at least being, amen, I don't know. But tell, but there had to be, you know, parents were probably involved here, but it had to be a big decision point for you What made you decide, I'm ready to go out into the world and do this thing and right now and, and, and leave school at that time?

I will say there's probably a long list of folks who also left Harvard and didn't do so well. So I don't think we should just—

No, no, no. That's, those are the only ones.

You know, survivorship bias here a little bit. I don't know if we, you know, there's obviously a lot of others too. Um, jokes aside though, I, I spent my time at Harvard. I was studying computer science and statistics. As well as a little bit of neuroscience. I was doing applied AI research. I met a lot of really smart people. And the way that I oftentimes tried to do that is I was running my own things to just meet people. I started running my own AI paper reading group. I started a startup group. I was just like, how do you get smart people in a room in a way that's low touch? Just bring people together. It doesn't need to be a super regimented club. I actually considered dropping out of Harvard to join a startup at one point and decided not to for a variety of reasons, but it boiled down a little bit to the people. It was like, who was I going to be working with? And that's actually what made this decision really easy for me. I started Sapien with two of my best friends from high school. Pranav was over at Stanford, Aria was over at UT Austin. They are two of the smartest people I've literally ever worked with. Pranav was one of the best infrastructure cybersecurity minds of our age group. He was hacking things since he was in middle school and got in a lot of trouble sometimes, but really did a lot of awesome stuff.

Any good stories on the hacking? Nah, I'm just kidding.

Not for public disclosure.

I figured you would say that.

He would make me go and scrub it from the record here. Um, and so, so he was someone I knew extensively and, and Aria also, like, he's awesome. He was a top 100 golfer in the world growing up and then went into computer science somehow instead of golf. Um, he ended up at a PhD research lab at UT Austin. He was their youngest, uh, student there. He was on track to get his PhD by like 22. So like two of the smartest people I've ever worked with. And To me, it all, it's about when you're building something, you wanna be building it with the right people. And so it was clear we had like the right talent density to at least get something off the ground and start building it. And then the other two pieces that kind of layered in afterwards is we met the folks over at, at Neo, um, Ali Portovi in particular, who runs Neo, the, the VC firm, and some of the other folks there as well. And it was one of those moments where you realize like the people aren't just the people in your company, it's the people you're gonna be working with outside. And there was that moment when they first made us an offer to You know, fund our pre-seed and kind of get the company off the ground. It was like, this is, these are people we wanna be more like, like these are just people who are really excited about working with. So I was like, okay, I have the right people to build this thing with. We then really like started to nail down this space of like, look, financial data is such a big gap. We actually initially weren't certain exactly where we were gonna take this, but it was so clear it's a humongous problem. And so to me it was like, look, I'm working with great people. We're gonna work on a really hard problem that has high impact. And frankly, the world is moving so, so fast right now. Like the opportunity cost of not going and building something is so high. It made it a no-brainer decision to just go all in and do it. And, you know, now we're almost 2 years into the process from when we first started the company. And I mean, it's, it's pretty awesome. Like, you know, team of more than 20 people now growing quickly. We're actually moving down the street to a new office in New York. Um, it was definitely a hard decision, at least for my parents was relatively easier for my co-founders a little bit of a harder time, but. I think when you boil it all down, I think life is being around really talented people and working on really hard problems. And when we got our first customer, started to really see real impact, it's one of those things that made it so obvious, like, why would we ever want to go back to school when you can be, you can be, you know, saving thousands of jobs, you can be generating tens of millions of dollars for these companies, you can be having a really big impact. Um, but that's kind of how it all came together.

Yeah, that's huge. And that's, I mean, and it sounds like, I think you're spot on with when you have those connections and, and those people that you have an opportunity to work with at a time like this is, you know, this is 1995 all over again. Well, probably 1995 on steroids.

Right.

Hopefully a better ending.

Yeah, but that's true. Well, I mean, you know, there's argue, I mean, there is obviously we're in some kind of bubble around it, but, um, you know, that also companies that started, you know, the Amazons that come out. So the best companies are gonna shake out and by starting building when you did, you know, you're getting, you have more, more of a foothold to, to ride this wave. So yeah, totally makes sense. And I guess my, my second part, it's not even the second part of the question. It's a completely different question, but it's, I, listeners to the show know that I've spent 2 years railing against everyone calling their chatbot with some different kind of wrapper on it. And Paul knows where I'm going, calling it an agent. And this year, We are actually, you know, there've been huge leaps in how long the conditional loop can go out and work without a human in the loop there. Tell me about when you first started building this, the idea of agents, how you were handling that, how much was left to the agent to kind of run on its own versus sort of an agentic workflow where you have deterministic steps in and, I'd love to hear your insights on how much better the large language models have gotten and the AI tools at going out and performing these long run truly agentic tasks.

It's an amazing question. There's, there's a lot that's changed here. I'd say candidly, like our, our approach is very different than a lot of other companies in the space. Like there's been FP&A tooling, there's been analytics tooling, there's been all these things for, for decades now, right? Our whole angle was one where. We were building an AI research team that really cared about financial and operational problems rather than, you know, an FP&A platform that was adding in AI. And what that meant is from day one, we were building the system we called the company— we called it the company representation. It's now kind of called the company engine. It's more, you know, user-facing, if you will. But it's this whole idea of like, why was this problem hard? Like, why did we have planning softwares that took 9 months to migrate into? Like, why were we even in this place in the first place? in the first place. It's because these systems didn't understand the business they were working with. And this will get to your question in a second, but it's really about, for every company, how do you build the underlying model of the business? And this isn't like a traditional Excel model. This is what are the parts of the company? How do they relate to one another? What matters to this business? How does all the data, all these different pieces together? So the whole idea when we first started was just, it was just this piece. It was, if we can understand a business well, then we could build the most general company out there, right? We could work with a manufacturing company and a software company and a restaurant and a telecom company and healthcare business, and they could all leverage this one platform. So that's actually where we started. We started with this foundation of understanding of business. What that let us do initially is actually exactly what you mentioned. It was just automating repeatable processes. The first workflows we did for businesses, frankly, it wasn't even an agent yet. We were doing it manually in Excel in the background, just like trying to figure out like, okay, how do you put this together?

Mm-hmm.

Um, that was where we started and it turned into like, kind of like you mentioned, these structured tool calls that went end to end. But the, the interesting thing about it is we were always building a system where it was extremely open and you added in constraints. And so what that meant is as language models got better and as technology has been improving, we just kind of loosen more of the little dials and the constraints around it such that you let this company engine do more and more and more of the work. As you kind of models, models get better. What this has led to is when we started, the use case you were doing is like, you know, we were making waterfall charts for companies based on their GL data, but it was like very methodical. It was like, there's only one way to kind of get there. We then kind of went to Sapien generating entire management reports end to end for companies, still reporting, but much more in depth, a lot more thoughts there. And the immediate value is. Well, you're going from a monthly surface level management report to can you get a mid-month, can you get a daily management report of the business that's going and breaking down to the customer level, to the SKU level, all those levels of detail. Then where we've seen Sapien go as these models have gotten better is again, loosening those restrictions. Now Sapien is actually going into things like pricing recommendations, inventory allocation, like the actual, like, okay, we've gone from saving you time to a system that is saving our clients millions of dollars or generating them millions of dollars where they're losing opportunities. And that's the big picture vision for Sapien is because we built the technology in this way, we're not just building a chatbot, we're building this like system of action, the system of data understanding, the system of frankly what matters to a business. Sapien is able to then go and get proactive. And so where we've seen the most value today is we have companies that feed in every single piece of their data into Sapien. It's their ERP, their warehouse, their spreadsheets, their CRM, their HRIS, all of them are flowing into Sapien. And Zapier is sitting there and it knows what matters to the business and it's saying, well, I know you're trying to expand your profitability in, in these segments of the business. Here's actually what happened in the last 24 hours here, and here's what you should do about it. And you're moving like less and less of needing the human to guide every piece of it. But what ends up doing is now every person is way more valuable. As we talked about at the beginning, now your people are not sitting there doing that manual spreadsheet stitching. They've moved up and up and up and up and up the stack to now it's strategic decision-making, executing on those decisions, going and making the actual changes in the business. And we've seen a massive shift both in how good models are, but also a lot of the work is outside of these language models. Cause frankly, LLMs shouldn't be doing most of this work, right? They're not built to do math. They're not built to do repeatable calculations. So it's all about how do you augment them with the right pieces. And I think we've kind of had the right timing where as LLMs are getting better, we can loosen up some of the pieces, but also. our own proprietary technology, like our tabular reasoning model, which builds understanding of every column and row and table, our calculation engine, which does all the math. They all kind of ride this wave together where they're all getting better. So there's a lot of different pieces there, but I think it's been really interesting to see. We made this really big bet when we started that people do not just want a faster Excel or a better Excel. People want— A system that helps them run their company better. And ideally you can also, you can export to Excel and keep some of the same ideas. That has definitely started to pay off as models have gotten better and better and better because this was not really even possible 18 months ago. But now we're seeing companies that are optimizing their business straight on a platform like this one.

You already had the guardrails and the systems in place, so maybe, but I imagine, like you said, as the technology gets better, yeah, it, it enhances what you've already built. But have you seen anything in the way that you're set up where you're able to turn over more to the truly to let the LLM do the agentic work, uh, than you were just a few months ago? Or do you feel like, and we still need to keep these guardrails on, so we're gonna, we're gonna keep it locked down and on it, not let it just go rogue on us?

Yeah, totally. I think, I mean, the way I would break it down is we definitely have less guardrails on the LLM than we used to in the past. Um, and I think it's one of those things where like you sometimes have to build stuff just to get rid of little bits and pieces. Pieces as the technology is improving. But we try to think about like, what are the parts that are— we call them like invariants. Like, what are the parts of the product that we're building that no matter how good an LLM gets, you're going to need them? One example is auditability. Every single number, you need to know exactly how it was calculated, all the way down to the source data, every assumption that went into it. And so we feel very comfortable spending time there. We feel very comfortable spending time on the data understanding layer and those sorts of pieces. One that we actually spend a lot of time on that has still proven to be really valuable is what we call financial intuitions. But it's just like the system having a gut feeling for like, hey, when you're comparing periods that aren't apples to apples, like how do you think about that? Like when you're creating a table or a graph, like what is the way that people just like information presented? This is something that you would think would not actually be necessary, but when you ask, you know, any LLM to go and build you a board deck, it's going to get all these little things wrong. Like they don't make a huge difference. There's also the bigger things going to get wrong, but all these little pieces have come together. So I'd say we've let the LLMs be a lot more flexible, but we, we do see a lot of places where the work that we've done to build up these guardrails is still very, very valuable. And, you know, companies do their 10-Q reporting through Sapien. They do like processes that we would've never imagined being done through the platform. And that's only possible because you have these financial operational guardrails that make it structured enough. And I'd say that's actually the big weakness of LLMs in general is they're almost too open. Like you can ask anything, they can give you anything. There's no governance, there's no control, there's no structure. Companies get a lot of unique value from Sapien because you're building a dashboard in one team and you can share it to the other team and they can go and drill in and get an understanding of it, but you still keep all the right, you know, structures in place. Um, but we've definitely seen it get more and more open, I'd say over, especially over the last, you know, 4 or 5 months. There's definitely been a big shift with the last few models. But there's definitely still a very healthy tension between the two that our research team spends a lot of time thinking about.

I'm gonna stop, Paul. I could keep going, but we're—

I know you can.

I can go on and on.

I'm gonna let you talk.

I'm looking at your face, Paul. I wanna ask more. You know, one thing when you said, you know, kind of the guardrails and they've got a lot better, but an LLM can kind of do everything. It made me think a little bit of what's been the, what's made Excel so incredibly popular. But it also has been the bane to many people is it's incredibly flexible, right? You can do anything in Excel. You can do anything with an LLM. There's a reason I like to joke, and I think, you know, in some ways this is true with an LLM, a little bit different, but, you know, Excel is the number 2 software for pretty much every category in the world.

Right, exactly.

Right? You know, it's not number 1 for anything, but it's always gonna be cheap. And if you know enough about Excel, you can pretty much build it to be almost anything, at least at a small scale. You know, you could do pretty much anything with an LLM, but does that mean it's really the best tool to use? And so I think there's some similarities there in what you said. So anyway, that just kind of, kind of came to mind.

There's one thing that I actually, I tell our team very often, which is I think in this day and age, the best products are opinionated enough to give you immediate value, but flexible enough to let you take it the rest of the way. And what that looks like at Sapien is We onboard companies really, really fast because agents do most of the work to go and understand the data. It's like, you know, multi-billion dollar companies get onboarded in 2 days and it's like, it's really powerful what this actually looks like. The first thing they get when they open the platform is not an empty chat box. It's, hey, CPN's already gone and it's calibrated on your business. It's thought about your company. It's researched your competitors. It understands your space, but it also structurally understands your data. And you know what? We already found 3 SKUs that are gross margin negative, and we found these customers that you should be renegotiating with. That's the first piece of it. So it already understands that it's, it's thought about it. That's the magic moment. But then everything is configurable. You can go and take it that extra mile and go and ask whatever question you want, but it's structured enough where it's gonna still give you all the pieces of data that it took to get there, like all the pieces together. But I think this is actually exactly what I, I find the issue with Excel, with LLMs, et cetera, to be is, are they putting the power in the hands of the user? Yeah. It's tricky to tell. They're multipliers, right? It's how good is the person that using this, and then you can multiply them. I think the best pieces of AI technology are ones where it can— anybody can get value out of them, but the power users can get even more. And I think that's like a lot of what we think about and why these opinions matter a lot, is can you build that piece of software? Like, we have, we have non-finance users use the platform a lot— operations, commercial, sales, etc. They're all using the same data, and it's a big boost for these folks to not have to go and ask their IT team every time they want to pull some data. But that means there's so much consideration for like, well, Sapien can't just show you a 300-line SQL query and say, do you think this looks good or not? They have, they have no idea. Right. So there's so many things like that where, you know, it's that, it's that careful balance of like, do whatever you want with like, can you have enough of an opinion to not just be number 2? I actually really like the way that you framed it. It's something we think about a lot here, both on an engineering side, but also on the you know, sales and how do we enable customer side of things.

Ever feel like your go-to-market teams and finance speak different languages? This misalignment is a breeding ground for failure, impairing 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, B2B revenue. Qflow 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 stack, makes segmented scenario planning a breeze, and closes the planning loop. Create airtight alignment, improve decision latency, and ensure accountability across the team. Appreciate that. One thing I wanted to bring up that I think is interesting, I think Glenn will find interesting. Now, you and I were chatting earlier with the agent tools. We've seen this with some other agent tools. But I know you specifically, you're not storing any of this data. So right now, right now it's not going into a data warehouse. You're storing kind of the metadata, some of the schema. Maybe talk a little bit of why that thinking kind of benefit. I think that's so new to many people to think, okay, it understands all my data, but it's not storing it, right? That's not traditionally how it's worked.

It's part of why the company's called Sapien, right? It's like, what's the human way of doing things when you're doing analysis? You're not taking the entire Excel and putting it in your brain and jumbling it around. You're looking at bits and pieces of it. When you're working with a data warehouse, you're not pulling every single piece of data warehouse. You're just pulling the pieces that you need. What this does is two really great things. One is, again, it makes setup really easy. We don't have to go and do a big ETL transformation, a consulting project, you know, these long ways of getting data into a system. It's working with it the same way that humans do. Yeah. And so that means is when it's doing analysis, all it needs to know is what data do I have access to? Like the names of the tables and the descriptions of the tables. And we generate a lot of other pieces that make it— CPM very effective at searching over this data. But once it's searching over the data and then it has to go and make a good query to pull in what it needs, there's two options. Either you go and pull in tens of millions, hundreds of millions of transactions, or you make a system that's smart enough to know what do I need to actually answer this question well? And that is a huge shift in efficiency, right? Questions that used to take days to pull the data for, if you just queried it right, you could solve those problems much faster, right? We see even today, some of the finance teams we work with, they previously were asking their IT team to go pull like 6 months of transaction data to go do an investigation. And that might take the IT team a couple weeks to, you know, get on their plane and a few days to do it or whatever it might be. So if you can go and do that full loop, it's going straight to the source data, it's grabbing the right relevant source to do the analysis, doing it, and then it's spitting it back out. And what that means is also it helps out with security. You know, a lot of folks are very nervous about security broadly right now. Is AI training on my data? Is data going outside of my four walls? And there are still bits and pieces where obviously we need to work with the data to do the analysis, but it helps that Sapien doesn't have to just store humongous copies of all of the information within companies. And I think especially as we're seeing more and more sensitivity to data security, this has been a really big selling point for us. It's like, look, we can do it fast and we can do it much more securely than systems have been in the past. You don't have to trust us as much as you might have if our system operated differently. Um, and it also just yields better results. Like this whole method means that when there's data in NetSuite and Salesforce and Snowflake, Sapien can go and quickly scan across all of them and see, oh, there's conflicting data between these two, or there's potentially, you know, there's a revenue column here and a revenue column here and a revenue column here. Okay, I'm going to actually go pull all 3 of them and compare them instead of having to go and look one table at a time, which, I mean, also we think about context windows, right? That's a big limiting factor for these AI systems in general. This helps avoid that problem entirely where you're just bringing in what you need and then the non-LLM querying and calculation systems can do everything else that you need to do.

It sounds like you're trying to take away my day job, which is, believe it or not, I'm not a professional podcast host, but doing these, uh, sort of Yeah.

Sorry, Glenn, you have two podcasts, that put you at professional status.

Everything that you just talked about is right now, if you don't have a piece of software like Sapien, you hire a consultant like me to come in and build an MCP server or some sort of connecting fabric so that your Snowflake data and your Salesforce data and your NetSuite data can talk to each other and be aggregated. And it's, it's like the new gold rush that there was around building data lakes and data warehouses and all that back in the day. But seriously, the great equalizer of AI is, you know, you with a sort of a foundation as a statistics guy will appreciate this, but, you know, we've all known, anyone who's been in the business world has understood conceptually what data science is and what machine learning is, but the barrier to entry is you have to be able to, you know, you have to write algorithms, you have to know Python and have the, data science fundamentals down. But now when you can create all this with just natural language prompts, it really lowers the barrier. It also leads to people doing really stupid things because they don't have a statistics background and they don't know things like you alluded to earlier, like survivorship bias. But to that end, and I think that a point with Sapien and, something you mentioned earlier that all the best AI platforms are easily configurable. And I think there's a move right now, you know, to some people Make and Zapier were easy because they were drag and drop and, you know, you had to sort of have the process flow. But now when you're setting up work in Cowork, for example, or you're, you know, there's other products out there that will sort of automate your workflows. And I think even in N8N, you can go in kind of type what you want and it'll build sort of the skeleton of what you wanna build. But you also, both you guys were talking about, you know, Excel is, you can do all kinds of things with Excel. People know different levels of, you know, you come in and you're like, yeah, I can put in an A in this column and then I can put what A means in the next column or whatever. And then a few years later you're using Lambda functions and let and whatever anchor array and all the,, Excel skills that Paul knows way better than I do. But I mean, where are we right now in the world with, yes, you want everybody to be able to do this. I keep thinking of AI in so many ways as like giving individuals their own RPA. Like, it's like now, okay, instead of having to have a big vendor come in and set all this up, everybody can kind of automate their own, but they can also get into trouble. So I'm wondering where— And maybe it's really on the engineer of the, on the software designers. You want to give them that flexibility, make it configurable, but also keep them out of trouble. Where's that line? And what do you, like, what's your dream for the, you know, how humans interact with Sapien and how they're building this and using it? There's two questions in there.

Yeah, I know, I know that's it.

But I hate questions in that.

I told you, Ron. I know, this is, look, there's nothing better I'd like to do with my time. My time than dig into these questions. I think about these all the time. So we'll take part 1 as like, how do you think about the extent of configurability? And then we'll talk about part 2 more as like, where does this then take us? Um, I think this goes to a big reason why I think like something like Sapien is so valuable. It's because the line for configurability is different for every use case, right? It's different for software engineers versus for finance teams versus for sales teams. Like it's just, it's fundamentally a different problem. In fact, even Sapien, like one of the things of configurability is like Sapien is a very— we actually didn't start at all thinking about forecasting, but we ended up building out an incredibly robust forecasting system. Why? Because Sapien had every piece of financial and operational data to build the best possible driver-based forecast you could ever imagine and have AI agents go and reason about every line item and build out this forecast. Now you could go ask like an LLM to build you a forecast. But there's so many opinions that went into it. Like, oh, if you want to break down a row, how do you do it? If you want to, you know, extend the timeframe for this forecast, like which models should it use? Like all these different questions. Like think about like when you tell an LLM build a forecast, there's literally 50,000 little decisions that have to get made to make that forecast go right. And that's the theme broadly across the platform. So the question is, how do you solve this? And honestly, like this is, I think one of the hardest problems in AI is what's a good assumption versus a bad assumption, right? If I go and I type into Sapien, what's my revenue? It's a good assumption that I'm talking about my company, but it's maybe not a good assumption that I'm talking about 2024 or 2025 or 2026. So how do you know which things you should just guess and which things you shouldn't just guess? A lot of it is a hard engineering and research problem that depends on the context. It depends on the company. For example, if we work with a windshield manufacturer and we work with a paper manufacturer, which are two of our clients, If one of them asks about, you know, how should I think about our cost structure? Well, one of them cares about sand and one of them cares about wood and like completely different ways you'd have to think about that. So this all goes a little bit back to that. When it comes to configurability, a big piece within Sapien is continual learning. So as you use Sapien, it's building up a knowledge base. It's building up to understanding of the data, but the human is always within control. So as we work with enterprises at larger scale, there can be like thousands of pieces of knowledge that Sapien is learning about the company. And you wanna have some way to control and govern that, right? You don't want any person to be able to go make changes, but you also don't want it to be a chore. Like you don't want someone to click and hit approve on like every single piece of information. There's been a lot of work on like, how do you let the finance team iterate and, and make changes and let the operations team do their own thing separately? And how do you let it all cascade up? Um, and I think this goes a bit to the finance intuition piece that I mentioned earlier. It's like, When you ask Sapient for a waterfall chart, it should know what a good waterfall chart looks like versus a bad waterfall chart. It should know, you know, when you're asking for the P&L, like how do you actually break out the P&L for this company? This all comes from context. And so I think like this configurability problem, it's a really, really hard one. But if you have the right context and if you have a healthy governance structure that actually lets you do a really healthy job of configurability because context makes sure that the configurability makes sense. And governance to make sure that the configurability is controllable. And once you have those two things, and then goes to how we think about outputs, there's no such thing as 100% accuracy. And this is always like the, the sentence that gets people really scared, but it's like, look, if you ask your analyst tomorrow, how do we compete with our biggest competitor in California? There's no answer to that question. Like there's no right answer. There are so many different ways to go about it. This is where verifiability matters a lot. And so we built the platform in a way where every number, every assumption, every calculation is traceable. And that makes it such that when you're doing these configurations, you can see where they're going wrong very easily, and you can go and make those changes. So it's kind of, there's so many pieces that go into it, but if you have context, flexibility, governance, and verifiability, we see that being a very healthy combination. And all 4 of those from the lens of the finance and operational use case. It's very different than what people want for their coding agents. Like for coding agents, like show me every line of code. That's great. For Sapien, we have users who have no idea what SQL even is using it. And we have people who have spent their last 10 years writing SQL queries. How do you make one system that works for all of those? It's hard. And actually even part of it is Sapien can actually treat people differently. So like if a CFO is on the platform versus a head of FP&A versus a head of IT versus an analyst, they can actually tweak Sapien to approach them in different ways. Like how deep should the analysis go? How technical should the analysis be? Be. So it's also kind of meeting users where they're at. I wanna then jump to the second piece of the question, which is like the, where do we go with this? This is really the part of Sabian that I'm most excited about. I think, again, this is where I think we're very different than other companies in the space is we're obsessed with this idea of can AI run companies? Like, can it do all of the financial and operational, like grueling analysis in the background and just like free up people to do the work that we're meant to do? Yeah. And that's different than, hey, we're building a system to do faster reporting, or we're building a system to answer your questions. We're building a system that needs to understand the company as well as any person does, and arguably bring in all the little bits and pieces from everyone. The long-term vision for where Sapien sits, it's, it's these agents on top of your ERP, your data warehouse, your CRM, your spreadsheets, all these different systems. And it's just telling you what you should be thinking about all the time. 'Cause any business knows there are 10,000 things you could pick from. But if you had a system that understood enough about your company, was configurable enough to know what you care about, and it could serve us up every morning, what are the 5 things you should dig into? That's the most valuable thing to pretty much any company out there. Then now the next piece is, can it actually go and act on it? Can it go and help you? That's where, again, maybe human in the loop, you don't want it to go and make changes to the business without having some control over it. But that's the future that we're excited about is you just explain to Sapien what matters to you as a business. Sapient helps you figure it out and helps you get there. And where it thinks you're thinking about the wrong thing, it can also disagree with you. I think that's one of the hardest things in AI also is it always just agrees with people. But if you can grout enough of these financial intuitions, enough of what matters to the company, it can be like, hey, I know you're asking me to improve gross margin on this set of your SKUs, but your real problem's over here. Right? That's what you want the system to be able to do. And that only happens from all those pieces getting layered together. Um, So that's a lot of what we think about. And it's been very exciting as LLMs and as Sapien has gotten better to see it get closer and closer to this vision where Sapien really is sitting on top of, you know, we're in public companies where it's sitting on top of like tens of millions, hundreds of millions of transactions, like 8 different systems in a way where people have never looked across these systems together before. They've never married the data in their on-prem data, like data warehouse with their Salesforce. They've never even connected the two together. Now they can, and they're suddenly realizing like, we're just losing, we're losing money on like half of our customers. And it's this realization of like, wow, like there's so much to do here. Um, so a lot of stuff there, but I think it's, it's been a really exciting part of the journey, certainly.

Yeah. And I love to hear all that because there's, you know, there's sort of the hype debate around, and, and I do agree with actually with Yann LeCun around this, that, you know, oh, the LLMs will never get you to AGI. For people in business who are using this, I don't need AGI. If you can have a system that can go across all these disparate data sources and find these correlations and actually have the context of my business to give me value, I don't, I mean, it's the same thing I say about working with Paul. I don't care if you're dumb as a brick, if you're a good co-host. Oh wait, did I say that? I kid, Paul, I kid.

You know, what did I hear once? 70% of, uh, all truth is said in humor.

Uh-oh.

I like to say we're solving the most obvious problem in the world. Um, and it's just about, can we solve it in a unique way that nobody's done before? And that was what we, like, from day one, like when we were starting this, like when we first pitched this to CFOs, it was like, what are you talking about? It was like, it was like, this is so far-fetched. And so we kind of, we pared it back a little bit and we went, look, Well, we can automate any workflow in your financial and operational process. Like, tell us what you're thinking about right away. But it quickly came back to this whole vision of like, yeah, this is the most obvious problem in the world. Companies are losing money. Companies are missing out on opportunities. It's because people don't have the time. One of our, one of our manufacturing clients, like they're based, um, they have like a plant in, in Tennessee where they now go and dig in to see like, hey, there was a college football game last weekend. Like, did our production drop off? That's a question that nobody would've asked before, but now they know, oh, people come in hungover the day after, or the Monday after they, they skip work. Well, now we know about it, and now we know that we actually should staff a little bit more in the days after big college football games, which like— That flavor of question, people never would've even tried to ask it cause it would've taken them weeks to dig into the data. But because it takes a couple minutes, the whole script gets flipped. Like companies are just thinking about every single point in their business they can optimize. And I think it's, it's really cool to see what effects that can have.

All right, so we're going to move into our AI section. How this works—

Sorry, were we not already in the AI section?

No, this is the, uh, AI question section. So we fed your, uh, profile, the questions we asked, your bio, you know, the web. We turned an LLM loose and said, give us a little bit of kind of personal, maybe a little quirky questions to ask. So we never know what we're gonna get. I haven't read them.

I'm scared.

I'm not even sure I remember which LLM generated it. I think it might have been Copilot on this one. We move around.

Well, that's concerning.

Yeah, we'll see. You might be able to tell. We'll see. But, uh, here's how it works. You got 2 options. I'm kinder than Glenn when I do it. You can pick a number between 1 and 25, so we'll put the human in the loop. Or the random number generator can pick a number between 1 and 25, and that's the question I read you.

Fair enough. Um, 13 is always my favorite number, so we'll do that.

Lucky 13. All right, this is on the— on building and running the company. You're in your early 20s running a company scaling fast with executives from OpenAI, Google, and Microsoft on your cap table. What's the hardest part of managing people and advisors who are decades more experienced than you? Yeah, did a pretty good job. I— we get some, we get some duds, but that one's pretty good. The golden retriever in accounting was a dud.

That was a weird one. That was—

I actually really like this question because, look, I, I spend a huge amount of my time building this company. Like, it's every waking minute. It's, it's not a lot of sleeping minutes, frankly, that exist. Like, And the part that's really fulfilling about it is, of course, building really cool technology and, and having this impact on our customers. But the other side of it is leading and aligning a group of people is probably one of the most powerful things you can do. Like, one of the hardest and the most fulfilling tasks, especially for us. Like, talent density is a, is a really key thing. Like, every person at Sapien is someone who is incredible, who I would wanna work for in another life, who, like, maybe, maybe one day I will work for them. Who knows, right? Like, That's the type of bar that we have. That means that organizing that team and keeping everybody in line, as well as like working with incredibly experienced investors and customers, is, is a big challenge. I think it all boils down to like, I'm a very first principles thinker and a lot of our team kind of approaches this as well, where I try to just think through like, what makes sense? It's not, oh, we should go do this sales process because some other company did it that way.

It's—

Okay, well, why did it work for them and would it also work for us?

Right?

We shouldn't just go and copy, you know, Facebook's engineering philosophy. They built that engineering philosophy 15 years ago. There's probably good nuggets in there that we should be thinking about, but what actually makes sense for Sapient and who we want to be and how do we bring those pieces together? And I think that's what oftentimes, like when I have to work with people who might be far, far more experienced than I am, it's a good way to level set. It's a good way to kind of bring everybody to the same table because it's no longer about You know, I now am someone who really appreciates someone who's been there and done that and brings a lot of that thought process behind it. But I'm also always trying to figure out, like, to the point on where we started, survivorship bias. People can give you really great advice on how they did things and they were very successful, but do you actually always know if that is the main reason why they were successful, or is it just a small part of it, or is it a luck, or is it other pieces that come together? Um, and so as a team, we just try to have a culture where we encourage disagreement all the time. I always tell people, like, just because I'm the CEO of this company, like, I'm, I can be wrong often. And I, I, you should be calling me out. You should be disagreeing with what I'm saying. And we just have a very open culture where between team members, between team members and investors, between team members and customers, we have that culture where, you know, we say always assume good faith. The person on the other side of the table from you is, you know, you have to assume you're trying to work on this thing together. You're trying to get to somewhere productive. And if you have that culture, then it's very easy for someone might be 10 years more experienced or 20 years more experienced, but you're sitting at the same table trying to solve the same goal. It should never really matter who has more experience. It should matter who's bringing the best idea to the table and is it actually gonna solve the problem? And it's a big call. Like we have a whiteboard over there in the office that just like the number one thing is always assume good faith. And I think it's a value that once we live by it, it makes it very easy. Like someone could be double somebody else's age, but everybody here just knows the other people are here because they're incredible. Yeah. And everybody's gonna work together to solve these problems. We have like engineers working together with people who spent 10 years in private equity to go deploy Sapien at a, at a PE portfolio company. And sometimes the engineers will say, I think we should do it this way. And, or the person who's been in private equity will actually go and like, we have internal coding tools now. They're like, you know, building out features in the product and telling the engineer, no, you should do it this way. Like that's the type of culture that we try to create. And it makes it a lot easier, like as, as we grow and you know, part of our team is very young and part of our team is more experienced. I think it helps really like balance all of those pieces out together.

Paul, which, uh, which LLM created these questions this week? Because there's, there's like some really good ones in here, you know, compared to something else. I'm gonna have to look.

I honestly don't remember because it was the day when I did like 4 of them and I kept switching between different LLMs. It may have been one of them that day I know was Claude 4.7, one was 365 ChatGPT, and I think Or not ChatGPT, Copilot. And I may have used ChatGPT for the other.

I don't remember. I think I'm gonna call an audible here because, so what I normally do, Ron, is instead of your random number generator, I just dump the questions back into an LLM and say, which one of these questions should I ask? But today I'm jumping in, I'm stepping in as the human in the loop because in the bottom under the personal and quirky section, first off, we've got Paul Barnhurst. We've got Chelsea and Real Madrid. We've got Spikeball, which side note, nothing ever made me feel more uncoordinated in my life than trying to play Spikeball. It's an old man problem, I think.

Wow.

You really just—

My son— Every piece of information about me online, it seems.

Yeah. But the one that I'm most—

Yeah.

The one that I'm most interested in, and that, you know, this is like a weird, like another opportunity to talk about myself, Paul Barnhurst. Paul, and I'm gonna— I want to talk about the 3D printed fidget spinners. But the reason that was interesting to me is I was not a great student in, in middle school. Um, I was too busy thinking I was really funny. But, um, for some reason, whatever version of economics we had to take in middle school, like, we had to start this business and sell something, and we had to just pick some sort of crappy baubles. And, um, I found some like acrylic keychains that people could put photos in, but I turned them into rock and roll keychains and I cut, this is how old I am, I cut pictures out of magazines. Have you ever seen an actual printed magazine? I cut pictures out of magazines and I had my Led Zeppelin and my The Doors and whatever keychains I was making and I made more money than any other business. So there's that story about me and we're gonna finally get back to what we're supposed to do is ask questions of our guest.

That's all right, Glenn. We can humor you. Like I said, you know, senior citizen. And all.

So you grew up selling 3D printed fidget spinners. And this is actually, I'm reading this from what the AI spit out. So I don't know where this was located, but it says, you grew up selling 3D printed fidget spinners and running a rubber band bracelet operation where you employed other middle schoolers for a cut. I just want to hear about it.

I think it was third grade. It was elementary school. Okay.

It would be great if you were in third grade and had middle schoolers working for you. That would be, you know, maybe they could have been the—

That could have been my answer to the last question is I've been preparing since I was in third grade.

So I just, and there's more to the question, but I just honestly, I don't care about the rest of the question. I just want to hear about that.

I think I've always like, from a young age, I was somebody, I like to say, like, I just had a bias towards action. Like, I was just like, if I couldn't sit still, like, I just wanted to always go do something. My grandfather, like, ran a furniture shop. And I think that was one of the things where I saw like this idea of like, okay, you can like, you can just create something of your own and you can just put it out there. So yeah, growing up, like, in 3rd grade, I got like one of those Rainbow Loom kits that like everybody was Making, and I, for some reason, had to take it to the extreme of, well, I was gonna start making these bracelets and then I made a website and I started like putting them on there and like trying to sell them. I then actually started selling them in school and I basically told, I got, was getting too much demand. So I started telling my classmates, I was like, well, you guys also have this. Do you wanna make some money? Like if you make one, I'm gonna sell it for $4 and I'll give you $2. And I made a few hundred dollars off of this within a couple of weeks and the school got wind of it and they were not particularly happy. They actually ended up making me donate all of the money that I made. But in some ways, I actually think that was like kind of a cool ending too, because I realized like, okay, I can do something really fun. I can learn a lot from it. And I could actually put the money to a good use because frankly, I was not going to do anything useful with it. I was probably going to go try and like buy an Xbox or something. It was a cool experience. In middle school, I was like 3D printing fidget spinners, kind of the same idea. I got to keep the money that time. I actually started printing them on the one 3D printer in our school. So I was printing them on there. And then eventually, like, I was like, like monopolizing that 3D printer too much. So I had made a few hundred dollars at that point. I went and I bought my own 3D printer and just started printing it at home. Like when I was at school and I would just bring in a new batch of like, you know, 10 or 15 spinners every day. And went on Amazon. I ordered a bunch of like the bearings that like go into those spinners. And there's like a huge box on our front porch one day. And my mom was just like, what is going on? Like, How do these even get here? I think I was always just somebody who like, I felt like I wanted, like if I saw something, I wanted to just, I can't even explain why. I just always wanted to do more. That's, I think why I ended up dropping outta college to start a company. I think it's, I mean, we didn't even touch on like, I spent a few years in high school like building a liquid fueled rocket, which was like the next escapade. Like I was on a goat farm pouring concrete and like building electrical systems. It's like there's so much to do in the world. I've always just wanted to like, if something's in front of me, I wanna, my friends will probably tell you like, I do something either 0% or 110%. I have no understanding of what doing something 50% or 80% of the way there looks like. Like there's two extremes and I find myself on that 110% extreme pretty often. And probably a good thing to be a startup founder in that way.

This totally could have turned into, you know, one of those 4-hour podcasts. Uh, I guess we're supposed to cap this. I think we're already long, aren't we?

We are.

We are long.

I have a daughter that will be waiting for me at school.

So let's not, let's not, let's not do that.

She'll wait a couple minutes, she'll be okay. I'll just tell her it's Ron and, and Glenn's fault and she'll be like, who? So honestly, I think I could go a couple hours, but thank you so much for joining us, Ron. This is a lot of fun, really enjoyed chatting with you. Great question to end on, Glenn. Well done. Continue to be Miami Vice. And until next time, thank you everybody for joining us.

Appreciate you both.

Thank you. 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.