The Exchanges

Q And is it, is it like a replacement ROI that you're selling? Like you don't need air specialists anymore, or do you like enhance, like what's the positioning?

A Yeah, it kind of just depends on the company. So, you know, we sell to flyover state businesses. I'm from Michigan. So think like industrial manufacturing distribution, those types of companies. And we really augment teams. Some people are like, you know, we might want to move accounts receivable back to sales. You know, sometimes there's people in like warehouses that do it as a part-time job, or even like our larger enterprises like Honeywell, they just don't have the account penetration. You know, think like somebody in accounts receivable can probably cover like 200 accounts a month. Uh, with AI, now you're able to cover like 5000. And so, and most companies actually are losing about five percent of their revenue trying to collect payments and matching up all the cash and handling with all the bullshit.

AI assessment note: “And we really augment teams.”

Q So maybe to kind of clarify, like, we just had Shahar from Terra Security on here. They're kind of like AI pen testing sort of thing, right? At least that's how I understood it. Is that, would that be similar to what you do, or not so much?

A Uh, yeah, we do a lot of that. We, we view like automated penetration testing as like a use case of our platform. And generally what happens when you have companies that focus on that single use case, they're probably going to make something that's more broadly and widely applicable. Maybe it does like a better job at certain compliance things. You can offer it to like SMB and lower size customers. It's very PLG friendly. We are Building something more big and interconnected. So although there might be capability overlap, there's probably practically like no customer overlap. We're targeting Fortune 500 USG, which includes Department of War and like the federal government and state government. And they require something that's a little bit bigger, more intense, a little bit more trustworthy and something that is more cross-functional in nature. So on the surface, probably competitive, but like practically no. And we think that there's a big enough market to support both

AI assessment note: “So on the surface, probably competitive, but like practically no.”

Q Oh man, that's crazy. So we'll get into the story here at LifeKid, which, which started about five years ago, I guess in twenty-twenty-one, but maybe a little bit of background, because I know you had some other startups before that. So maybe tell us a bit about kind of Your history and maybe a little, you know, even more about that. Exactly the previous startup, right? EV Labs.

A Yeah. Um, so this is my fifth company. You know, the first one that's really kind of started to do well, but, uh, I, I grew up always wanting to start a company. My dad was in tech startups in the eighties and nineties during semiconductors and GPUs and DSL and all of these kind of foundational technology shifts, mostly on the hardware side. And, you know, I got out of school, and I'm like, I'm gonna start a company. So I, I went and I joined, um, Y Combinator, the 2007 class of Y Combinator. So it's the fifth batch. Did a company then, and, you know, I was a kid right out of school, and actually most of the founders at that time were just kids out of school.

AI assessment note: “this is my fifth company. You know, the first one that's really kind of”

Q see it, and that almost always leaves this, like, overhang of preempt, of like, your likelihood to get preempted just based on that, even if your investors are the same, and your, you know, delivery is the same, just the fact that you raised a bunch of money, it's like, oh my god, how do I, you know, this must be hot, how do I get into the next round?

A Super well said. Yeah, there's a, uh, frenzy dynamic, right, where you're like, oh, it's oversubscribed, I didn't get in, I gotta get in at the A, I gotta get in at the B. Okay, so we looked at this data. And what it showed is that on a percentage of graduation basis, yes, the highest valued companies graduate to Series A a little bit more often than the other companies. But it's not a massive difference. It's within a couple percentage points. The companies that didn't had a big, big change were the lowest valued companies. So the lowest quartile, if you were raising a seed round and it's in the zero to call it 25th percentile of valuation, so on the very low end, those companies made it to series A about half as frequently as the other companies.

AI assessment note: “the highest valued companies graduate to Series A a little bit more often”

Q This company started in 2017, but I'm sure the idea came before that. Maybe give me just your background first. Like what were you doing before you started, uh, Exonia's?

A Sure. So in short about me, I'm originally from Israel and I had a pretty challenging childhood growing up, uh, both economically, socially, and that's what made me very entrepreneurial as a person. Cause I always knew that I had to earn everything I wanted out of life and had to be really independent. And I really loved computers, right? Like, I learned how to program from a book from the library when I was 12. I was part of the team that won the International Robotic Olympics in South Korea when I was 15, and I finished my bachelor's degree when I was 19, then went into unit 8200, which is the signals intelligence unit of Israel, very famous for it.

AI assessment note: “finished my bachelor's degree when I was 19, then went into unit 8200”

Q preventable? What's not preventable? I unfortunately had a friend of mine whose dad just passed away recently from a heart attack, super healthy, you know, ate well, you know, exercise, late fifties, like, you know, completely inexplicable. Like, is that, do you see that like somewhat commonly, or is that really like an odd case? I mean, I, you know, I don't know anything about this world, so really curious.

A Yeah, I'll try to sum up some of the relevant statistics, um, you know, in efficient way, but, um, heart attacks are the number one cause of death. Uh, they're the number one cause of death in men and women. Cardiovascular disease deaths are the number one cause of death and, uh, total 40% more deaths than all cancers combined. Um, so just a massive thing. In the U.S., there'll be somebody who suffers a heart attack every 40 seconds. Uh, which is like why we all know somebody like your friend's dad, right? Who, and then the most common presentation for having heart disease like that is not presenting with chest pain or shortness of breath, but it's simply having a catastrophic event, dying of a heart attack or suffering a heart attack at home with no antecedent warning. And that's why it, it's a silent killer in the majority of people. And the most common diagnosis is suffering a catastrophic event. So We know how to treat somebody. Once we understand that they have heart disease, we have plenty of good treatments for it. Like, um, 90% of heart attacks are preventable. Uh, we, it's not that we lack the treatments. It's that we're just not identifying people soon enough so that we can prevent these, um, events from happening. And the last statistic I'll give you is the average age of sudden coronary death, like fatal death due to heart attack in the U S is about.

AI assessment note: “the most common presentation... is simply having a catastrophic event, dying of a heart attack”

Q And what was the YC experience like? I mean, so many people obviously want to get into, into YC, like, give me a sense of what it's really like to kind of go through that, how helpful it is. What are the sort of things that happen?

A Totally. So we joined for two purposes. One was, of course, the brand recognition that goes really far away. Actually, at least three businesses. The second was, hey, we're a first time founder. There's probably a ton of things that we don't know. And so having this like trusted advisor who Can help us like navigate all those, you know, plot holes. Uh, let's do that. And then the third part was, uh, you know, it was a, it, it could be a tremendous distribution channel. Like you hear all the time about these YSE founders, like buying and selling from one another. Now, out of the three, the first two were correct. Great brand recognition. We got great advice from our group partners. Uh, distribution, uh, not at all, you know. Like we did not sell to other companies in our cohort, which a lot of YC founders do, uh, you know, because their ICP tended to be more later stage. And, you know, if you reach out to a YC founder of like a series C startup, I mean, every batch, I'm sure they're getting spammed by YC founders trying to sell them stuff. And, you know, you're not getting as much responses.

AI assessment note: “Great brand recognition. We got great advice from our group partners. Uh, distribution, uh, not at all”

Q What happens in between like that 2003 and 2005 when you really started?

A So I lived in Johannesburg, uh, for all of 2004. It was supposed to be a one week trip that turned into a two week trip that turned into a three month stay that turned into, why don't you stay there for the full year? And I love that time. I got to help establish the Endeavor office in Johannesburg, South Africa. Endeavor had only been operating in Latin America up until that point. And so this was sort of the proving that in the Endeavor model of supporting high growth entrepreneurs around the world, wasn't just a Latin America phenomenon, but could be applicable broadly, globally. And I got to work with incredible mentors, many of whom are still mentors and role models today. Foremost amongst those was a guy named David Frankel and David Frankel, um, had heard the Endeavor case study as a business school student at Harvard business school. He was a very successful South African, uh, entrepreneur before going to Harvard Business School, heard the Endeavor case study, was heading back to South Africa and said to Linda Rotenberg, the founder of Endeavor, you've got to do this in sub-Saharan Africa. This model is great for Latin America, but you should really bring it to sub-Saharan Africa, make that your next region. And this kicked off David becoming really the founding board member of Endeavor South Africa, helping us to establish, uh, an office, a board, Helping to hire early…

AI assessment note: “I lived in Johannesburg, uh, for all of 2004.”

Q hundred, like that, that part's not, I mean, with angels is actually, that's maybe, maybe I'll ask that even like, if you're doing an angel round, like what, what have you found in terms of getting in front? Cause like VCs have websites, you know, angels typically don't these days, like where are you seeing founders either meeting or like sourcing enough angels to have like enough of these conversations?

A Well, at the risk of being, uh, you know, self-promotional, obviously we have a database of that. We've got about a 130,000 angels in founder suite. I'd say number two, looking at LinkedIn, you go in and type the words angel investor into the search bar on LinkedIn. It's like over a 100,000 results. And then, you know, narrow that down by, by maybe your geography and second degree connections, things like that. Right. And, you know, Usually when you do that exercise, you'll come up with a list of maybe 2000 or so angels on LinkedIn and then go in and look at their profiles and again, qualify them, look, look for clues that they do your deal. So LinkedIn's pretty good. Depending on where you live, you know, angel groups can be good right here in the Bay Area. I mean, we have, there's probably a dozen different angel groups. There's life science angels, right? There's all these different, um, different angel groups. Now that's a, that's a whole different process because you usually have to apply and get someone to vet you and Vouch for you, and it can take a while, but still, you can hit, you know, 60 angels in a room or on a Zoom call by working angel, angel groups and angel networks. Sometimes also looking at similar companies going in, plugging them into Crunchbase, seeing who funded them. You'll identify some names. Also keeping an eye on, uh, I call the industry, you know, t…

AI assessment note: “obviously we have a database of that... number two, looking at LinkedIn”

Q Dude, so you just raised just over fifty million dollars from Andreessen, you know, one of the most well-known VCs in the world. And you're doing like, you know, AI agents in healthcare, which I think these days, everybody's doing AI agents, except you've been doing this since like, 2018, 2019, you know, five, six years ago. How did you kind of get into this world so early on?

A Before that, I helped start Google's AI venture fund, Gradient Ventures, right after the attention is all you need paper came out. So the thesis there was, the world is going to change because of these breakthroughs across every major industry, and how can Google invest in that growth, both from its own platform, but also innovations around the ecosystem. And so I had a front row seat to seeing how this technology was being used in many different ways, And in one time, I was showing my wife a demo of a project that was automating phone calls to make spa and restaurant reservations, and she said, Ankit, incredible technology, good demo, but why do you need this to make spa reservations? I wish someone would bring this to healthcare. She spent her career in healthcare, and she said, if you could make information and knowledge available to people when they need it, how they need it, if you could unlock time that was scarce in healthcare, where Data is exchanging hands on phone calls. It would change how healthcare in the U S works.

AI assessment note: “Before that, I helped start Google's AI venture fund, Gradient Ventures”

Q Is it like a Costco credit card, like that kind of a thing?

A That's a co-brand, but yeah, same, same idea. So it's a financial product that is offered through a brand, typically a retailer, but, um, obviously airline cards are huge, does also fall into this bucket. Those products create tremendous value for the brands that offer them, uh, the customers who adopt them and then ultimately the issuers as well. Um, so it's a really interesting model in that sense for the brands they're, they're building a deeper customer relationship. Uh, it's a much stickier customer relationship. Oftentimes they're earning a revenue stream out of the financing or the, the, uh, the financial product. Um, and you can go look at, you know, these public companies, like they are significant revenue streams. I think Target gets a check from their issuer each quarter that's in the neighborhood of like one hundred and fifty million dollars. So we're not talking about small potatoes. It's, it's like real money. Um, for the customers who adopt, if it's a brand that you love and you shop there a lot, It unlocks the best rewards. So typically at least five percent back, there's all sorts of perks. And then finally for the issuer, you're getting to originate lending accounts to customers with effectively a zero dollar cost to acquire, which in consumer lending is like the holy grail. So there, it's a really, it's a great business model. It's been around for decades, bu…

AI assessment note: “That's a co-brand, but yeah, same, same idea.”

Q So there's overhead plus a fee you're paying to this bank for them to sponsor you? Okay.

A Correct. Because they're effectively the true lender in this, in this model. And so, uh, Pablo shows up at Home Field Apparel, which is one of our partner brands. You want to buy some, like, college gear. You originate an account. We do the underwriting. Uh, so that's part of like the Capital One background comes into play there, but we had to design our own credit policy. Once we approve you and give you a credit line, you make the purchase. Uh, Celtic funds that loan. So we settle with the merchant every day. So we're, we're sending ACH out every day to the merchant. Um, a couple of days later, we buy the loan from Celtic. So there's then a cash transfer from us to Celtic To effectively acquire the receivable. We then take that receivable and we park it in our warehouse facility. Now, in an ideal world, you could park the entire receivable there. Um, but in reality, the way this works is there's typically what's called an advance rate. And so let's say your advance rate is 90%. Uh, the warehouse facility is going to fund 90% of the loan. You need to fund 10% of the loan. And that's your capital.

AI assessment note: “Correct. Because they're effectively the true lender in this, in this model.”

Q Were there any, like, major in those first 1224 months, any kind of major celebrities you landed that you remember, like, being kind of an inflection point?

A There kept being different inflection points. Like, the first major one was when Cody Ko joined. Cody Ko is one of the biggest YouTubers on earth. Yeah. Also happened to be one of my fraternity brothers and me and Devin, my co-founders roommate. And at the time when Cody joined, we were just doing, uh, NFL players basically. And we started with, you know, pro athletes only. And then one day Devin's like, I think Cody and people like Cody might do well on cameo. And when Cody joined, he put it in a YouTube on his YouTube channel. He probably had a few million followers at the time. And that was like, Like, just the first time the site went viral when Cody came on. Um, you know, shortly after that, uh, Dennis Rodman came on, which, like, was the first talent that really was, like, press worthy. This was probably, like, a year and a half, kind of, into the business, and I remember, like, the Chicago Tribune was the first paper right about Cameo, and the headline was, for 200 dollars, Dennis Rodman will wish you happy birthday, and, like, that was bold and intriguing, and him coming on, Uh, ended up being big, and then probably right after that, like, Brett Favre is one of those people that just, you know, this guy is a Hall of Famer, and someone that was, like, so popular in pop culture, and, and just, like, when a Brett Favre's on, it's hard for any athlete at the time to say, li…

AI assessment note: “the first major one was when Cody Ko joined... Dennis Rodman came on... Brett Favre”

Q And by the way, what was the, Like for them, what was the story to them? I mean, obviously there's a financial piece where they think they're gonna, you know, earn a return, but like strategically, were they thinking they're gonna be partners? Were they thinking what exactly?

A So definitely, you know, if you're looking at companies like Hewlett-Packard, Cisco, Hitachi, they would be selling the hardware platforms. Hitachi and Hewlett-Packard, they're also storage vendors. If you're looking at the component vendors like Microne, Seagate, Western Digital, Samsung, they want to understand from the forefront what would be needed in a few years. How do they create differentiated products? If you're looking at companies like Mellanox or Nvidia or Qualcomm, they're a strong ecosystem. Obviously Nvidia invested before they became such a great data center company, but they, they had the strong hunch that We would be making a big, a big difference in the whole story. Mellanox invested because a part of the reason we could take what we're doing to market is their fast networking. So all of these, these companies look at these investments as, as a way of peeking into the future. They all have their CTO offices that have ideas, but investing in startups help them validate. Uh, they all want to see Which one of them are winning?

AI assessment note: “all of these, these companies look at these investments as, as a way of peeking”

Q I'm always like, when you start a startup, your goal is for that startup to work, to get really big, to be successful. And so I'm always shocked, like when that happens, and then the founder's like, let's do it again. You know, so like, well, what kind of was happening? What mindset were you in at the time? And why did you, why did you end up doing that?

A Mainly it was driven by I wanted to have a different kind of impact on a society and building a business, helping digital marketers, driving more traffic, improving the ROI on digital span is important, extremely good, but still you are few hops away from impacting people directly in a positive fashion. And I started thinking about what can I do in the second part of my career, my life, that will be a lot more impactful for the society. And start, after exploring education and healthcare, what I realized is that employment is the single most important thing in our society. If I can help people get a better job, better employment, I will have a much bigger, more fundamental impact. And that was the reason why I left Bloomreach to start Eightfold.

AI assessment note: “Mainly it was driven by I wanted to have a different kind of impact”

Q And so moving on to seed, then what's, what's a typical seed round?

A So seed can happen either on safes or in price equity. It's kind of fifty-fifty muddy middle now. We only include seed. We only call around a seed round on safes if it's for two million or more dollars. One of the advantages of being the system of record here is that we can split out really easily Primaries versus bridges and extensions. So because we see that actual information, we know if it, if you'd already raised the seed that this is a seed two, a seed plus, et cetera, and we can remove those from the medians. So it's not getting too messy. So proceed on Carta today. The median amount raised is about three to 3.5 million. And the median valuation, this is a pre-money valuation is 14. So if you add those up, that's, you know, call it seventeen million post or so. 17 to eighteen million.

AI assessment note: “The median amount raised is about three to 3.5 million.”

Q you think like what that makes me think about, like the other obvious test is like NEPR motor score, right? NPS. And it's like, how, how likely would you be to refer this to someone? Like, why do you think that, The Sean Ellis test is a better representation of PMF than, than NPS. Like what do you think one gets at that the other one does and vice versa?

A I love NPS. I think NPS is a, is a great, uh, survey as well, but I think it's more of an operational, it's like kind of an operations feedback. Like let's say I have a product that's a must have, but their customer service is terrible. Like the person yelled at me in customer service. I'm probably not going to recommend that product to someone else, even though I personally need it. I just hate doing business with that company. The, the likelihood of recommending the product is more a function of your entire experience using that product where, um, from a product market fit perspective, the, the first most important thing is do people actually need the product in the first place? Great customer service on a product that no one needs isn't going to get you very far.

AI assessment note: “first most important thing is do people actually need the product in the first place?”

Q What was that? Tell me just quickly, like the, you told me how it would play out, but is there a valuation implied or like what's, what's the structure of that agreement?

A Uh, we used a pre-money of four million and a post of Three. So our investors own a little over 25% of the business, um, or sorry, right around 25% of the business. And, uh, we have, uh, essentially they own units in an LLC. Although a friend of mine just fundraised for his business using the SparkToro docks, uh, and did it with a C Corp. So it's very, it's at your option what you want to do. Um, both, by the way, qualify for QSBS, the, the qualified small business, uh, stock. Exemption, which means that the first ten million dollars of, you know, earnings that you make on a, on a sale of your units or sale of your stock is tax-free. Uh, so that's very compelling for a lot of investors too. Um, and the valuation is flexible, right? So I think that because of my reputation and network, my valuation was higher.

AI assessment note: “we used a pre-money of four million and a post of Three”

Q And how do you use this regularly in the business? Like, is it like every Monday morning, everybody sits there and looks like, you know what I mean? Like how, how do you build this into the process?

A So each team, right? So marketing team, sales team, value delivery, customer success product, they have their own agents where they're running the types of insights and the type of data that they want from this. Right? So for example, like one of the things that we really care about right now is driving the usage across our customers of our MCP. Connector because people who are using MCP are, are shipping more agents than people who are not. Right. So we really care about driving the MCP usage across our customer base. And so the product team that's working on MCP gets a report every day in the morning and says, this is how much usage there was. This is the sentiment. This is the feedback that the customers are giving. These are the types of queries, like all of these things. And so every single day they can just prioritize and iterate and making the product better. For the customer in the, in the right direction.

AI assessment note: “the product team that's working on MCP gets a report every day in the morning”

Q Tell me, this is a good, good time to segue into that, like, what exactly is Judy Health? What, what do you guys do? And, and even, I know the name has changed, I don't know if what you do has changed, but curious what it was at the beginning as well.

A So, Judy Health is a health benefit manager, so anyone who works for a company and you have a benefit plan, Somebody's administrating your benefits, so that's medical, pharmacy, and the longer tale of dental, vision, and ancillary benefits, but the point of it is your insurance or your benefit provider is the person that manages your account, you know, and so this is what Judy Health provides. We started in pharmacy benefits, and name of the company in the early days was Capital Rx. Talk about Gritty. People often say, oh, like, what was the name? Capital RX, what did it represent? I'm like, nothing. The domain name was available. Nobody had any trademark rights on capital RX. So it was clean. And more importantly, I wanted it because it sounded common and people are like, what? Because in healthcare, you want to feel like you've been around forever.

AI assessment note: “Judy Health is a health benefit manager... We started in pharmacy benefits”

Q What about, uh, in terms of finding these candidates, did you any, like, how did you, those first 30, 40 hires, did you do anything interesting on the outbound side in terms of finding them in the first place?

A Yeah, so, alright, first 40 employees, I go to one degree of separation. You know, is it someone I worked with, or someone very close to me that I respect worked with this person? So, I'll give you a great example. Uh, I was looking for someone on the implementation side, so this is someone who knows how to set up a benefit plan, which is extraordinarily difficult. And I asked all around, because, you know, I've been in healthcare for a while. I know people in different companies. I'm like, who's the best person in implementation you've ever worked with? And I kept getting the same name, this woman by the name of Karen Durker, who still works with me here. And I'm like, what the heck is a Karen Durker? And I'm like, what are the odds three different people both said the same name?

AI assessment note: “first 40 employees, I go to one degree of separation.”

Q And you started this with Anton as well? Same co-founder?

A Yeah. So I started with Anton that I started my first business with, and then I had a friend from 10 years back whose name was Gustav. Uh, and Gustav at the time was working at Goldman Sachs in Stockholm. He was part of a startup that got acquired by Goldman. So he was at Goldman building autonomous trading systems as a developer or engineer. So I called him, he was one of the smartest guys that I knew at And I said, like, we think we have something here. We showed him the actual workflows of how you, like, what are the state of the art in how you design and engineer technical systems to buildings? How does these processes work? And he was blown away. He asked us questions like, how did the global engineering community and the global engineering industry accept this as standard? Why haven't someone done anything about this before? And, um, those were sort of the right questions that, because those were the questions that we had as well. So we, he almost verified our, our thesis a little bit, but he said, I will join as a co-founder, but I have one requirement and that requirement is that I can bring on a fourth co-founder and his name is David. And my, my other requirement is that he's becoming CTO. So we got connected with David. We had never met David before, but he had been working with Gustav for in Goldman Sachs for, for many years.

AI assessment note: “Yeah. So I started with Anton that I started my first business with”

Q Where are you asking them specifically at that point?

A You know, at that point, it was less of a question, and it was more of a, hey, can we have a conversation? I would love to learn more about how you do your work. And so, in, in our specific example where we, in San Pablo, where we, there was a, a commander at the time, Commander Brian Bubar, who responded to our reach out, and we were essentially like, hey, we are really interested in this, this case that you solve. We were interested in learning more about how you solve these cases. Could you take us along as like a partner? I would love to learn more. And he said yes. He invited us to the police department. He gave us a bunch of background checks and basically handed us this book and he's like, all right, let's see what you can do. And we got to work and basically worked out of that department every day for 18 months.

AI assessment note: “it was more of a, hey, can we have a conversation?”

Q deployed for a month on a single customer. Let's walk through that. Like, tell me, you know, how you did it, why you did it. And like, how big was the services element? How do you still, obviously you ended up scaling. So I'm trying to understand all that, but let's start at the beginning. Just what was the idea around forward deployed engineers? How did you set it up?

A I think the overarching principle for us in the early days and still today really is over investing for success and Peregrine started with deep roots in product development and implementation. And that's what we really wanted to be good at. And we weren't so good at marketing and we weren't so good at these other functions, but those are two things where like, we're going to put all our points into this. And we really believe in those early days that over investing in, in ensuring the customer successful would pay massive Dividends. And I think that was also the, especially the case in government where it's such a highly high trust network. I mentioned that we got our second customer because the chief just told the other chief that these guys were good and trustworthy. That dynamic is very, very real. And so, you know, we weren't really thinking too deeply about how do we build an infinitely scalable business? Uh, we were thinking about how do we make our customer Really successful. And so even in those early days, I remember one of our first engineering hires, we sent him to our second customer in Pittsburgh, where he basically worked out of that office for about three weeks integrating their data. They had this like luau and he cooked a key lime pie and he just completely got embedded with that department, understood their lingo and was able to deliver a much better integrati…

AI assessment note: “we sent him to our second customer in Pittsburgh, where he basically worked out”

Q So back to the main storyline, you're the solo founder and you got the thesis, you've jumped fully on board on this. What's, what's kind of step one?

A Yeah. Step one, customer discovery. If there is one takeaway for anybody who's listening, if you haven't read it, read a book called The Mom Test. It is, in my opinion, the best book on customer discovery that is out there, followed by The Challenger Sale, but that's a sales book. Um, yeah, I kicked off what I called the 90 and 90, 90 customer interviews in 90 days. So I spent 90 days mom testing people. So not going to them saying, hey, my name is Damien. I've got this idea. This is what it is. Can you please tell me you like it, but rather going, Hey, I'm thinking about this space. Talk to me about how you've solved this problem. How would you rank this in order of priorities? Where would I focus on if we were to build something like this? And most importantly, friends, if you're doing customer discovery, always ask at the end, if they're willing to introduce you to somebody else in network, you know, because not only does that build a relationship with this person, but it means that you also get routed to. other like-minded folks, not for confirmation bias, but rather you're going to get to continue to see either in role or industry segment different folks. So it allowed me to build a pretty comprehensive market map of where am I solving this problem for whom and how would they expect us to do it? And to be honest with you, Pablo, it actually changed some of my thinking. It …

AI assessment note: “Step one, customer discovery.”

Q Was this for the first meetings or you did it only for the second meetings? How did you, how did you make this happen?

A It varies. So one of the key things that I benefited from was there was a core industry conference about two months into the 90 and 90 called black hat that happened for us. And I just like on my own dime, like flew out to black hat, got myself a hotel room, paid for my Ubers. And I just like walked the floor. I like walked up to strangers and I was like, Hey, I'm Damien. Can I get you a coffee? But it was putting myself in those positions to be successful. So it was for plenty of first meetings, and yes, definitely for second meetings. We had a bunch of folks, I'm based in Boston, Massachusetts. We had a bunch of folks in Boston, and it was like, hey, this was great. Hey, do you want to go for a run? You want to grab a coffee? And then I'd get out there and meet them where they were in their respective city.

AI assessment note: “it was for plenty of first meetings, and yes, definitely for second meetings.”

Q What's it, what's an example of some of those ABC, it's usually not that's like a revenue milestone, I assume. So what might that be?

A If you want to make a magic box that pulls clean water from the air, let's say just using sunlight, which we did by the way, and talked about on the podcast, then let's agree after a year of this, you're still going to have a super rough prototype. So you will not have scaled, but you need to be able to tell a Fairly exciting and specific story about how we're going to get down to a penny a liter, because at 10 cents a liter, it's glamping. It's kind of cool, but it won't really change the world at a penny a liter. You will change the lives of more than a billion people on planet earth. Super exciting. Can we agree that within a year you need to be able to tell that story about what the techno economic slope is like, and you have the next year to go figure all of that out.

AI assessment note: “you need to be able to tell that story about what the techno economic slope”

Q And so how do you go from, so that that's actually super helpful to understand how you land into this space of accounting. How do you go from the space to the idea? Do you then do like classic customer discovery? Like how do you figure out what this first product is going to be?

A The lesson that we learned early at Stripe was the best way to build stuff. And ironically, the easiest is find a customer that's willing to trust you with a hard problem. And if you solve that problem, they'll give you more problems. Hmm. It's as simple as that. And it sounds very, very simple. Obviously there's a lot of work into that, but if you're able to do that, you can repeat that motion over and over and over again and build a very, very successful business. And so we went, talked to accountants, looked at what they do, shadowed a bunch of them, talked to owners. We understood kind of the broader picture issues that they have with the workforce and everything. But in terms of like which space to start off with was very much driven by A lot of customers and these large firms asking us about it. Our first two customers were H&R Block and Armanino. And very fortunate to be able to work with those kind of tier of companies. Armanino, top-twenty accounting firm, some of the wealthiest individuals in the world and a bunch of tech businesses that work with them. H&R Block, huge amount of volume. Basically about 20% of Americans do their taxes on H&R Block, to give you a sense. Very different shape. Like the, the complexity, the domain complexity for H&R blogs, very different. Like the people that do their taxes, generally much simpler use cases than like kind of the wealthy al…

AI assessment note: “we went, talked to accountants, looked at what they do, shadowed a bunch”

Q you would go to try to build that company. Not to say it's easy to build the company, but the steps are clear. When you start how you started, which is top-down, so I've seen other founders do that. What I always want to know is, how do you go from there to the first thing? Like you mentioned fintech. How did you narrow all that to fintech, for example?

A Okay, a few things, but it's messy, so forgive me. So I started my career at DeShaw, which is like a hedge fund, and my co-founder went to Penn. So we had some roots, and people would, people will respond on LinkedIn if I message selfishly, because we have fancy LinkedIn's. B, many companies are massive in FinTech. Like, if you can touch the flow of money, I'm not going to recite the examples. You can build a very big business. The best case outcome is you can somehow index global payments, which is what we're aiming to do. And C is just It goes back to the second thesis. It was an incredible use case for AI. There's so much unstructured data. It is, like, purpose-built to use the models. And then we had a number of ideas. Some of them were derivative ideas of what's in the market already today. Some of them were quite boring. Some of them were innovations. And I made a fancy Figma as if we have them all. And I went to people, like, I was like, hey, Pablo, like, we're building this company. It's called Atlas. We're called Atlas at the time. And I already built these five products. Which one do you want to pay for? Kind of asked it as if we have it, like made a very fancy landing page, made each slide as if I have the solution and just was like, which one of these do you want to buy? I have them all.

AI assessment note: “I made a fancy Figma... and just was like, which one of these do you want to buy?”

Q about the origin story. Now that we know that moment when things started to find click, there was a year before that where presumably it wasn't clicking yet. Right. But maybe just start with that, like just a little bit of not, not maybe your full background, but right before you started, Amigo, what were you doing and why did you decide to start this business in the first place?

A Yeah. So I started my career at Google, spent time there, and then went to a company called Upwork. I was at the internal incubator there building sort of net user to one product surface area, sort of like new bets for the business, new ventures. And obviously a lot of the focus Upwork, like World's largest labor marketplace. We started thinking through what does it mean to be able to deliver high quality sort of knowledge work through the form of agents, right? So a lot of my sort of thesis and initial thinking was in that space. I left to start Amigo, which initially the thesis was similar in the sense of, is there really high value labor or knowledge work that is super expensive, super hard to find, supply constrained, That we can build and train these agents for that can then go deliver that quality of work, right? So it wasn't healthcare only in the beginning. That was the more generalized thesis.

AI assessment note: “I went to a company called Upwork... I left to start Amigo”