Pedro Franceschi: Lessons from an Evolving Entrepreneur, from Hacking iPhones to Building Brex

0:00Never disconnect yourself from the work that you put in the hands of the customer. Yeah, you're going to manage more people. Things are going to get more complex. You're going to get more offices and more locations and more segments and more products. But I think a lot of the way in which I learned to manage the company is just reviewing the work, never forgetting that this is what matters versus all the other things that get in the way. That's Pedro Franceschi, co-founder and CEO of Brex. At just 13 years of age, he was one of the first to jailbreak the iPhone. By 15, he had co-founded a payments company that would process over a billion in transactions.

0:35And in his early 20s, he helped transform a last-minute YC pivot into one of the most iconic fintech breakout stories of the decade. This is Mike Maples Jr. of Floodgate, and it's go time with Pedro Franceschi of Brex. This is Mike Maples Jr. and welcome to the Pattern Breakers podcast, where we explore why some founders radically change the future and how they stand apart. Together we'll learn about the counterintuitive mindsets and actions behind their remarkable success. Brace yourself for a world where chaos is welcome, naysayers are often a positive signal, and movements galvanize misfits who transform the impossible to the inevitable.

1:30Pedro's entrepreneurial journey has been quite a ride, powered by sharp technical instincts and an unapologetic ambition. But the lesson we can learn that may matter most is if you want to build something great, your rate of learning has to outpace the rate of change around you. Pedro and his co-founder, Henrique, didn't build Brex by having all the answers. They built it by constantly evolving. In our discussion, you will see how Pedro shows us that reinvention isn't just a one-time move. It's a habit. Adaptive flexibility isn't a personality trait. It's like a muscle that can be built with practice and commitment and time.

2:09And in today's startup landscape, where yesterday's playbook gets obsolete fast, It can be a vital founder superpower. Let's talk to him.

2:22All right, Pedro, welcome to the podcast. Thanks for having me. Before we even jump into things, maybe not everybody who's listening knows exactly what Brex does. What does Brex do and how's the product and set of services evolved over time? So at Brex, we are a spend management and financial services platform for businesses of all sizes, companies all the way from startups. One in three startups in the U.S. are in Brex today, all the way to the largest public companies on the planet. They use Brex to manage all their spend. What we do is we help companies spend smarter and move faster across credit cards, banking, spend management, bill pay, procurement, travel, et cetera.

3:06And you've had a relationship to the topic of fintech and also to the topic of just tech in general for a long time. If my memory serves me, you jailbroke the iPhone 3G in Brazil at age 13. How did that happen? And how did that sort of introduce you to the world of tech, if you will? Yeah, I was lucky enough to discover what I liked doing when I was very young. I started coding when I was nine, and I just had a lot of curiosity in how computers worked. And I was Googling and trying to learn English and coding at the same time, which is harder than it seems. And I had a lot of free time and ended up getting deep in this universe.

3:47And around that time, Apple launched the iPhone in the US, and I was using some tools online to jailbreak it so we could work in Brazil, in the Brazilian carriers. Back then, it was so the only way to AT &T. As part of that, I ended up learning how these programs worked. Then I built a new version of that that worked on a new iPhone that Apple launched, the iPhone 3G back in 2009. And, you know, I put it online and it became pretty known and sort of in this iPhone jailbreaking community. But I think the key idea is like, you know, the feeling of building for me came from a young age. And it's something that still is true today.

4:23I think the how has changed, right? I think I started coding and I still code today. but it evolved into company building and building teams and thinking about customers and going to market and all that. But building is, I guess, the core essence of who I am. And I've had it since a very, very young age. Thankfully, it was a super fortunate thing that I was able to find out super early in my life. And when you were jailbreaking the iPhone, were you sort of aware of the fact that you were kind of breaking some rules or was it more about, hey, I just want to have the empowerment that the iPhone offers, and that's the only way I can get it.

5:00What was going through your mind at the time? More of the latter, right? I mean, I think I realized that the iPhone was this super powerful device that was better than any phone that existed. But when I started jailbreaking the iPhone, there wasn't an app store. The app store didn't come out. And I was like, well, it would be really great if I could run my own software here. Really, it came from that sort of imagination of the possibility of what could this be if there wasn't as many constraints in the way that the product worked and I wanted to break them. Thankfully, jailbreaking wasn't breaking any patents or anything like that because you're doing it in your own device.

5:34But for sure, I don't think Apple was particularly excited about it. Did you get any communication from them or from anybody in there? Not from this one. I did actually make a proxy server that made Siri work in other languages back in 2011. And on that one, I ended up breaking some patents. So I got a notification, a legal notice from Apple and the company they acquired that was powering Siri. But I ended up shutting it down and it was just a really fun story back in the day. Yeah, I guess on some level, it's a badge of courage to get their attention. For sure. How did you end up meeting Henrique and how did that lead to regard out me?

6:16So like, I have to confess, Yes, I wasn't there at the time when you guys were building that. Totally. So we actually met through Twitter. Very random and funny story. We were discussing programming text editors. We just saw we had a lot in common. And I was working inside a payments company in Brazil. Actually, before that, my jailbreaking stints got this in the press in Brazil. And they needed someone that knew iPhone security back in 2011. There are not a lot of people that knew it in Brazil. So this company hired me when I was 14. And I went to work there initially to learn how payments worked.

6:53And so I saw from the inside how bad payments were. And then Enrique on the other side was using that on the customer side. And we had this idea of like, why don't we build a payment platform that's very oriented around developers? And we built that company for three and a half years from the two of us to, you know, hundreds of employees, a couple billion dollars a year in GMV and, you know, pretty profitable. And now the company is, I think, the third or fourth largest online-faring processing company in Brazil. So you built this company, and it's still in motion. I guess you sold it to Stone.

7:23That's right. And this is all before you applied to Stanford? That was around the same time, around the same time. So we sold the company in 2016. That is the same year we moved to the US and went to Stanford. We got in in 2014, and for two years, we just deferred our admission. So why go to Stanford? I mean, I think we always had this belief that the great companies were always built from the US. They were built from here, from the Bay Area. And for us, it was a lot about the people and the exposure that we were surrounding ourselves with. In hindsight, it turns out it worked out. But the company was pretty profitable, growing, and letting go of that to start from scratch again was not very obvious.

8:06Thank God we did it with the benefit of hindsight. but it was sort of giving up a whole life and something that was kind of working well to sort of give it a try in the U.S. So you're at Stanford and then you decide to apply to YC. Yeah. Originally with a very different idea from Brex. And I've heard the legend of this story from many people, including folks who are kind of your classmates or contemporaries. I think it started out as a VR idea. Yeah. So we had this belief that we started in fintech in Brazil. And when we moved to the US, we thought that fintech was a solved problem here. And I think what we learned is that there's this interesting correlation between how developed your economy is and how underdeveloped your payments infrastructure is.

8:55And when looking to Brazil, for example, in the 90s, you had hyperinflation. So the inflation, I think, got like 30 % a month. So if it took you one or two days to clear a wire or an ACH, you lost like 2 % or 3 % of the value. So a lot of what, you know, I was born in Brazil with real-time payments. The year I was born, real-time payments existed. And in the US now in 2025, you still can't settle an ACH in real time. Maybe a wire, but that's super expensive. And it's not exactly in real time. So it's been like a really interesting journey to just realize that the US actually was a much bigger market and perhaps even more underserved.

9:32And I think part of it too was, I think the more we spend time going deeper into where things were going and why they were the way they were, it became more clear that a lot of what we learned in Brazil could be pretty applicable in terms of building a company here. So then, but what happened at YC that caused you to say, I'm not sure I'd call it a pivot, right? Yeah. More of a complete reset change strategy. I mean, the first one is the notion that we couldn't really build VR with the knowledge that we had, right? So I think something that we really believe in with the benefit of hindsight as well is founder market fit.

10:13Like some ideas just require specific domain knowledge and honestly functional knowledge that we didn't have. And VR was clearly that. We knew nothing about optics, nothing about hardware. We were good software engineers, but that's where it ended. And then in fintech, we had the opposite effect. We had almost an unfair advantage to some degree because we've been doing it and thinking about it for so long. And one of the learnings that we had is like, look, any new idea from the outside looks very sexy and exciting just because you don't know the problems yet. The moment you know the problems, all ideas have challenges and things that are hard and, you know, something that requires a much higher bar in terms of how you think about executing at the highest level or at the frontier of the field.

10:55And in FinTech, we were just more equipped to do it. And then the second part that made us switch and sort of commit to it was we saw in our batch that, you know, 90 % of the founders didn't have a corporate card and couldn't get one. That was the important part. They applied for a card with Amex. They required a personal guarantee. And it was basically like putting their FICO and their personal credit scores on the line to run the business. And even when they did it, the limits were super low. So they couldn't run their corporate expenses there like servers and ads and all the things that startups typically spend money on.

11:27And we realized that there was a better way of doing it by looking at the companies and giving them credit based on hard data that is very different than what the banks were doing. And the initial data point we used was their cash balances because we thought that companies had raised money from investors. They were very asset light, so they were not financing inventory or anything like that. And really the idea there was because you can do this in a much more deliberate way when you have data, you can underwrite companies that would not be underwritten through traditional banks and give them a much higher limit.

12:02So we had this, you know, the value prop in the beginning of Rex was, you know, we were the first corporate card for startups with high limits, no personal guarantee and great rewards. And that was enough to get it off the ground and, you know, compound to quite a bit of revenue over the first two and a half, three years. And so what was the first demo like? I've heard stories about you had this hacked together demo, even with NYC that got people excited. How did that come about? Yeah. So early on, honestly, we focus a lot. The thing that we realized is the idea of Brex wasn't particularly new.

12:39People try to build this kind of product before. The reason they didn't succeed or haven't succeeded in the past was because they just couldn't get the car to work. They couldn't build a corporate car. They had to rely on these banks that existed for ages on their old systems. So one of the decisions we made early on is we said, look, this business is only possible to succeed if we build the core infrastructure from the ground up. So we spend the first nine, 12 months just building core financial rails, like authorization, clearing, settlement, underwriting, fraud, all these systems that people take for granted.

13:14The first 18 months of breaks were really focused on core financial service and just making the card work 100 % of the time. And we got to customer live in probably five or six months. But early on, that customer experience was very different than today. You had basically a command line interface. You had an NPM command that you installed in your terminal and you could access the card. Very old school. The thing that was interesting is the pain point that they had was they needed a card that worked 100 % of the time and that's what we were solving. So that's how I got off the ground. Over time, we hired great frontline people and put an amazing mobile and fast board experience.

13:53But that's not how it was in the very, very early days. Okay. So you had this sort of first command line demo that you improved over time. Did things start off like and just go up and to the right immediately once you locked onto the Brex idea? Or did you have any near-death experiences in the early days? so basically we we launched the card in june 2018 and the 2019 in probably 16 months we went from zero to 100 million ar so that was that was a lot of the sort of early days and then with it came a lot of the scaling issues and challenges of making a company work at at a very compounding path we we got a lot of we got a bunch of things wrong so first thing is we in 2019 i was 22 23 so we never obviously haven't really run a company at this scale before.

14:43And we went to 50 to 400 people very quickly. And a lot of the challenges of growing and hiring a lot of people sort of come with it. I think the learnings there was like, I think we were very focused on just serving one customer really well, which was like tech startups. That was the beginning of the company and really the inception of Rex. Because we were 22, 23, we had no option to hire people that were better than us. So we spent a lot of time trying to get these right folks in, especially at the leadership level. And we were super fortunate to hire an amazing CFO, an amazing head of sales, an amazing CTO.

15:20And all these things really set the company in a special path, I would say. But we got a lot of things wrong just because we haven't done it before. And I think teenagers writing code in my computer, and then all of a sudden the company was working and scaled a lot of people and customers and expectations. Right. And that was a that was a new thing to us. So, you know, it was a great year, but definitely tons of mistakes and learnings along the way. OK, so but was there a palpable moment where it was obvious to you? Oh, my gosh, this is working. I think I think the way it felt internally when it was working is a lot of stuff was breaking and yet growth was going up into the right.

16:00The product was broken. A lot of things are problematic and customers were upset. And yet they were excited and coming and, you know, like telling your friends about it. That's the biggest signal that how it felt like, I would say, you know, you're sort of the metrics are going off to the right place and people are excited about the product and so on. Yeah, it's sort of interesting. I find that when you have product market fit, there's still drama, but it's just a different kind of drama, right? It's drama that's born of success and people caring about the product and being upset at you because you're not doing the things they care about rather than just apathy.

16:38And how do I convince these people to want to care about this in the first place? But I find that there's always some type of drama. It's just a question of which type. So you started to take off. Was there sort of an initial set of investors or customers or banking partners that were early co-conspirators that sort of believed in what you were doing before the others? Who were the first true believers in Brex? Yeah, I mean, we had a lot of like the beauty of this model that we had is that a lot of companies started growing with us on Brex, right? So one that we're very proud of is Scale.ai. Scale.ai is literally number one customer on the database.

17:21Record number one is Scale.ai. Wow. You know, we went to deliver cards to Alex's office and it was him and a couple of people and a few engineers, his co-founder. And we gave him cards and they see his breast today, you know, seven years later. You know, the business is now, you know,$25 billion and so much compounding and now at the forefront of data and AI. So, you know, there's a lot of stories like that of companies who served super early. There's been, you know, of course, you know, banking partners that sort of abandoned us early on. You know, Emmerin Bank in the very beginning. Of course, a lot of the investors as well that believed in us before.

17:57You know, many other folks. YC did a phenomenal job in the Continuity Fund. Anu, the folks at Green Oaks have been fantastic. Folks at DST. Ribbit, of course, was a Series 8 lead. So a lot of the investors, I think, did a phenomenal job believing in putting their monies where their mouth was. And then MasterCard on the card side has been a pretty big partner as well since the beginning. I think there's been a lot of folks, but probably customers have been the biggest thing of just so many companies that sort of ran on breaks from day one and beyond. And even today, you think about so many of the companies that are at the forefront of AI, like Anthropic, Cursor.

18:39There's just so many of those that are fully running on Brax today and very excited that we're very proud and excited about it. You know, a lot of the startups that I've seen do really well. Sometimes they only realize it after the fact, but quite often there's a lot of inflections that created a set of external change events that contributed to their momentum. And so when I think of Brex, I think about in that window of time, you started to have cloud bookkeeping APIs. You started to have better KYC. You had better founder-friendly terms in terms of how companies were getting funded. You had more companies getting funded in the first place.

19:21Looking back on it, were there certain change events even bigger than you that you think contributed massively to your success? A couple of ones that come to mind. The first one is the fact that we could have the data available in real time, right? To underwrite a customer. So the invention of Plaid, they had competitors. Now they sort of dominate the entire space. But that was a big reason of why we were able to underwrite these companies. And that wasn't possible even two years before we started the company. So that was a pretty big one. Then, you know, I think as you project online, you know, obviously there were the demise of SVB was a huge sort of insane event for us, right?

20:00We got billions of dollars of inflow in less than 24 hours and thousands and thousands of applications and new customers coming on board in a matter of days. It was probably the craziest time running the company in the sense that it was just unprecedented and also dangerous because the whole financial system was pretty unstable for the course of four or five days. But a pretty fascinating time. and probably the third moment is just like, we're just seeing with AI now, so much of the buying criteria changing towards automation, which has been something that we've been preaching for a while. And two or three years ago, we used to go to customers and say, you know, you should care about automation.

20:42Like you should be buying automation. Automation is so important. It helps you free up time to move faster and go run your business and just do all these things that are more impactful than, you know, controlling your finances and your spend. And now it's, you know, it's the buying criteria in so many of the enterprise customers we serve and so many even smaller companies startups and maybe the last one i'll say is just companies becoming more global so we build like very deep product capabilities from a global perspective on brax like we issue local currency cards in 50 plus countries so if you have employees in more than one country you can always practice the same way uh and settle that locally with no effects and you know no accounting issues and and all that you know we we initially we build that for enterprise companies But we saw more and more that startups are becoming global sooner because they just have an employee in another country, maybe in Canada, maybe in Mexico, maybe in Europe, maybe in India.

21:34And all these folks can just use a seamless experience whenever they spend and, you know, move company funds around. You know, fintech is very gatekeeper heavy and you guys are pretty young. Were there ever times where you sort of had to bend the rules in certain ways? You know, you know, it's kind of like you hear about the Airbnb guys scraping Craigslist. where they're retired and not asked for permission. A lot of times. I think one of the things we learned and we run the company this way until today is that the best leaders, they have this attribute that we call the ability to operate at all levels.

22:11Being able to actually be a very high functioning IC in whatever function you're doing, right? So our CTO is our best engineer. Our CRO is our best salesperson. Our CFO is the best finance thinker in the company. And even early on, that was very present. Right. And one example of that is, you know, a lot of the times, you know, a very important attribute is just having this very clear ability to figure out things from first principles and distinguish like what is engineering and what's physics. Right. You know, engineering solutions for a problem. There's a lot of ways of going about it. There's a lot of ways to build a bridge.

22:50Right. But the laws of physics are the same. And a lot of what we talk about internally is like, what are the physics of a problem, right? And when we think about the way most people think about it, they always engage of the world and interact the world through the engineering concepts versus the physical concepts to keep the analogy alive. When we're thinking about Brex early on, and one of the things that we really wanted to bring live is this idea of instant onboarding. So when we launched Brex, the vast majority of customers went from zero to getting a card number that worked on an onboarding process in 10 minutes.

23:25No company on the planet had done that before, but you could never onboard, not even for consumer, by the way. You could go from zero to a card working that it can actually use and get credit and spend real money. The same way you sign up for an email account in 10 minutes online. I remember the first time we came to our banking partners and to our compliance teams and all the folks around the company that were very involved in the regulatory aspects. And they said, oh, this is illegal. You can't instantly onboard it. I said, you can't do that. That's just not OK. And the sort of obvious question that came in and asked was, look, where is that written?

24:04Where is that in the law? And they said, well, it's in the FinCEN requirements. Here's a number. Here's a lot. And a lot of what we did is we said, let's go five layers deeper. Let's actually look at what's written in the law, read the law ourselves, and a lot of the MasterCard manuals and visa manuals and understanding exactly the rules of what was actually mandatory. And it turns out they never say it's impossible to onboard a customer without a human reviewing it. What they say is you have to take a risk-based approach. It turns out that most of the times the way people implement that is they say a person can make a risk-based decision better than a machine.

24:38so you delegate it to a person. And therefore, people looking from the outside say, well, actually, a person's always required. But when you investigate the physics versus the engineering of a problem, you understand that that is the actual constraint, right? The constraint is like you need to have a risk-based approach. And the way we solve for that is you build a risk scoring model that in the beginning was very simple and now is very sophisticated that can actually determine the level of risk on an account better than a person now and actually make a decision on that before you even have a person reviewing.

25:10And then you only review like, you know, 10 % or 5 % of the accounts. And we did this in so many problems. This is just one example. But the way we build our banking product, the way we scale our capital markets, the way we build all of our AI functionality, and the way we think about automation, like we approach every problem with the same lens. And it just gives us a different edge in terms of how we go about thinking about the world. And so you rolled out Brex 3.0. What outcome proved that that change was worth the pain? And what was involved in that decision? And what were the sort of the trade-offs?