Y Combinator Startup Podcast - #82 - Henrique Dubugras of Brex and Anu Hariharan

Episode Date: June 27, 2018

Henrique Dubugras is the cofounder of Brex, which provides corporate cards for startups.Anu Hariharan is a Partner at YC.Brex went through YC in the Winter 2017 batch and just closed their Series B, which was led by YC Continuity.In addition to discussing how the Brex team built out their service, Anu and Henrique also cover the specifics of what it takes to build a fintech startup in 2018. And Henrique shares advice for young founders, as he started his first company at 16.Read the transcript on our blog.

Hey, how's it going? This is Craig Cannon, and you're listening to Y Combinators podcast. Today's episode is with Henrique Dubugras and Anu Hari Hurran. Anu's a partner here at YC, and Henrique is the co-founder of Brex. Brex provides corporate credit cards for startups, and they went through YC in the winner 2017 batch and just closed their series B, which was led by YC continuity. So in addition to discussing how the Brex team built out their service, Anu and Henrique also cover the specifics of what it takes to build a fintech startup in 2018. And Henrique shares advice for young founders because he started his first company at 16.

All right. Here we go. Hi, everyone. I'm Anu, a partner at YC continuity. And we have Enrique here, who is the founder of Brex, which was a YC Winter 17 company. The exciting news about Brex is they're actually launching soon and they offer corporate credit cards. They're basically redefining how corporate credit cards work for startups. So we're very excited to have Enrique here. Enrique, let's start a little bit about your background. We obviously at YC have

known you for a very long time, both you and Pedro. And you guys started your first fintech startup when you were at the ages of 15 and 16, which is not, you know, something we see every day. So can you tell us a little bit about what was the first company you founded and then how did you find your path to White Combinator. Yeah, for sure. So first, thanks for having me here. Yeah. And so Pedro and I, we both started coding pretty early.

I think that was super helpful. I think that was a little bit lucky. How early was that? So I was started when I was 12 because there was this game I wanted to play. It was like a paid game. And my parents didn't want to pay it for me. And I figured out if I learned how to code, I could play it for free. And Pedro, like, nobody knows how he started coding,

but he started coming when he was like nine years old. And then during our teenage years, we did a lot of like interesting things. So I participated in building this emulator for this game. And I learned a lot how to code because of that. And like I build like a version of that game. Pedro on the other hand, when he was 12, he found like the first jail break for the iPhone 3G, which got like a lot of attention. And so we we had doing our teenagers a bunch of.

of like experiences already in tech that led us to believe it's possible to start a company. So for me, it was that I was working at this startup and did the founder. I thought it was like a really cool guy. And my thought was like, if he can do it, I can do it too. And then I try to start an education company and that didn't work out. And then I met Pedro in the end of 2012 basically fighting on Twitter. programming text editors, Vim versus Emacs. And I was Vim, he was Emacs.

And I won. And we basically became super friends and decided to start a company together, which was Pagarmé. Pagarmé was like a payment processor, online payment processor down in Brazil. I think the comparables here would be like something like Stripe or Brinktree or Wepay or like those kind of companies. And we grew that company to a pretty reasonable size. It was like a really good experience. And you raised only $300,000. Yeah, it wasn't by option that much at that point.

So we were 15 and 16 when we started a company. So it's not like a lot of people in Brazil were willing to give us much more than that. And then after we raised the money, we were like, oh, this is so much money. Like, we can do whatever we want of this. And we were lucky that payments in Brazil. Brazil, it's in terms of margin a better business than most of the world. So we were able to have a profitable company since, I think, the end of the first year. Okay.

That's great. And that was super helpful because then we just grew with our own profits. We didn't have the need to fundraise. And profits solve a lot of problems if you can have them. That's true. And you sold the company four years later when your team size was around 100. Yeah, so we grew to a company, I feel the stats, but at the time we sold it, we're doing a little bit over $1.5 billion in transaction volume. We had, I think, a little bit over 100 people, I think 110, 115, something like that.

And we sold it September 16, so that was probably like three years and a half, almost four years. Not a lot of people have built 100 people team and they're less than 20. Why did you sell it? A couple reasons. I think that we always thought that Pagarmé could be a big business, but it couldn't be something, I think, as big as our ambition, because I don't think we had a lot of edge internationally. So Brazil is a pretty big market for payments.

You can see like Pag Sagurao just IPO for $9 billion, right? But I didn't think we had an edge to go international. We wanted to build a global company. And we got it into Stanford also. So we were kind of curious to see how that went. Stanford always gets you. Yeah, exactly. And like from international, it was my dream statement

was 14 to go to Stanford, right? So it was kind of like that. And I remember, you know, right from your application days, one year into Stanford, you guys had applied to YC with actually an augmented reality startup. Yeah.

And by the time you graduated from YC, that's how the, you know, you launched Brex. Or you were working on the Brex idea. I hadn't launched yet. Can you talk a little bit about, like, how that transition, happened. You know, you coming from AR as an idea and then saying, no, you know, we are going to work on the corporate redefining the corporate credit card experience. Yeah, for sure. So after we got

to Stanford, we're like, we're not going to do payments anymore. Fintech, too complicated. You know, if there's like old banks, you know, like that's, we're not going to do that anymore. Like, we've been through that. We're going to do something new, something easier. And then we thought AR was easier. Definitely not. And then we started thinking about these ideas, et cetera. And then it was actually three months into Stanford that we decided that we wanted to apply to YC. And we applied.

And we got into YC of this AR idea. And then a few weeks in and were like, okay, this is complicated. Like we don't know anything about hardware. We don't know anything about optics. Like we could only build the software. but like if Apple killed an API or Google killed an API, like then the business would be ruined, you know? Like there's all these things that we started to think about it. Then we're like, okay, that's not what we want to do.

So then we were like in this limbo and we're super worried as, oh shit, we took YC money. Should we give it back? Like is that something that we should do? And we talked to parties like, no, no, no, go figure out something. And they're like, okay. And then we started thinking about things we wanted to do. We went through like a bunch of ideas. And then we went back for an idea we thought for a long time.

We were like in Pagarmere, we were a payment system, right? So we received the money and then we paid out the merchants. And we're like, well, why do I need to pay out the merchants? Why can I just pay out the suppliers directly, right? Or the employees, why can I be like? And we started grasping this idea of, oh, what are the B2B financial services that we can generate value, right? And we actually thought about creating a bank. But creating a bank is very complicated as we went to find out in the U.S.

And then we were like, okay. Especially in the U.S. Yeah. So we went and said, okay, let's do a product of a bank. All these products are super good, right? And then we looked at credit cards that seemed like a one-of-a-product, but we didn't actually go too deep into it. And then, but we kind of had this landscape of like B2B financial products.

We were kind of looking. And then we wanted to get a corporate credit card like ourselves. And that was kind of it. And then we were talking to one YC partner, Dalton, who also said, hey, all these startups, they have a lot of trouble getting corporate credit cards. They're like, okay, let's investigate that. And then figured out that one, we couldn't get one. Why?

Because, well, because they wanted us, we went there and they're like, well, you don't have any financial history. I was like, but I have $120,000. Like, that's a lot of money. From Y, C. Yeah, from YC. like you can just give me $5,000 in limit. I'll be fine.

It's like, no, we won't give you it unless you personally guarantee. But we didn't have FICO score because we had just moved to the U.S. So we couldn't personally guarantee the card. And then we went to talk to like a lot of our matchmates and like a few of them were like, oh yeah, I can't get a card. No, none of the international ones could get a card. And the other ones, a lot of people chose not to get a card because like they didn't want to personally guarantee the card. Because as I agree with them, like the whole point of building a corporate.

is not having like personal liability attached to it, right? So there was all these people who are just using debit cards and like just walking around of debit cards of $120,000 in the bank or just using their personal cards and all these things. It's that, okay, that has to be inefficient. Like there has to be a better way to do this. And then that's kind of like how Brex came to be like by seeing that problem of our batchmates and ourselves not being able to get a corporate credit card or having to personally guarantee it. we had the idea of building something better. So let's talk about the team, right? One of, you know, you guys launched, you launched recently,

but you've been there for 18 months or now. And so how, when you launched Brex during YC and right after Demer Day, how big was the team? And can you also talk about your first 10 employees? Like, how did you hire them and how did you decide who they would be? Yeah, for sure. I think that one thing that we really believe, and I know everybody says that is that,

the employee, the 10 first employee is going to set a foundation for a later. And I think that we really took that to heart. When we got our first 10th, so in procurement, we didn't have a lot of options in their first 10. So we just got the smartest people who were, there was even this guy who didn't know how to code at all, but he was like a physics Olympics champion. Like, well, physics seems pretty hard.

Coding, I think is easier than physics. So we might as well just hire him and teach him how to code, right? So I hired like people that we knew that we were the smartest and we just like made it work. And Brex, I think we were a little bit more targeted. We were like, okay, those people were really good and they're doing really well now. But like if I could do it the right way from the beginning, what would I do? And our first two hires were actually like a CFO and a general counsel. Oh, wow.

That's quite, I would have never, I mean, you know, I know that. I know that now because we know you guys, but like I know most startups are not hiring a CFO and a G.C. Yeah, exactly. Because going of that, the thing about getting a really good partner and being able to rebuild everything from scratch and having like the good flow and all of that. Having a staff that like we're still young, right, we're 22 under 21. So it's not like we're older, but like not that old. So when we go to bank, having people who are more experienced than you are. and having them be the people that actually help you do the plan and describe everything

is really, really valuable. And we were very generous of the offers and the packages for these people because we really believe it could change the way the business went. So we got. And what I'm saying? Still, let me ask you this. It is very hard for a founder this early to be able to hire a really credible GC and a C. phone, right? No matter even if the offer or the package is attractive, because they, you know,

they probably, they probably have multiple options. So how did you get them over the line to believe in your mission and join Bregs? So I feel that I think this is a little bit of the advantage of being a second time founder is that I think people who are experienced, they want to work with other people who are experienced. So that helped, I think. But you guys came from Brazil. Yeah. They probably didn't know you. Yeah. So I think that basically, so usually founders, they have this, right?

Like they have this first product. And that may or may not have product market fit, but they have this thing. And then they have this like long-term vision. And they, that's like very, very far away. And they usually don't have the plan in between that much. I think these like more executive level hires, they want to understand better the plan in between. I think that's is something we had really well defined since early on. It's like, okay, we're starting with like this credit card for startups, which I think it's a great

market. And then we have this big long-term vision. How do we get from here to there? I think making that very clear and like how they could like how they would be the difference between getting there and not getting there also, like if they could add a lot of value there, was something that I think got them across the line. So I remember specifically of RGC, we were talking about how in order to be able to scale a lot, we would do a lot of the regular Tory things he didn't do in his last company. He would do it right this time here and how that would change like this, this part of the plan to actually work versus other companies that failed.

Like that got him really excited because, okay, I learned so much from the experience. Now I wanted the opportunity to make it right, to do it right. So I think that's how I got across the line. but also having a really strong set of investors helps a lot being able to get where they need on cash. Like a lot of times people want people to take huge pay cuts. We were never that type. We always was pretty generous on cash, like those kind of things. Got it.

And then so they were your third and fourth higher. Talk about this next six. So who were they and where did they come from? So we didn't have an easy time recruiting after that. Most companies don't. Yeah. Because we were from Brazil. So it's not like we knew a bunch of people.

People say, oh, what about your Stanford networkers? I was there for three months in my network. It's all freshmen. So it's not like I knew a ton of people. Yeah, you were going to be class of 2020. I remember. Exactly. I was going to be class of 2020.

So maybe in 2020, I have a bunch of people I can hire. But it was really tough for us. We basically started skewing towards people that were similar to us in some sense because it's what we could. So we got a bunch of international folks. Yeah. So I think our first engineer was Uruguayan and then, you know, we got a French dude and we imported someone from Brazil. I think we just did whatever we needed to do for the first 10.

But we didn't lower our, like we didn't have a quality bar that was lower because of that. We still had like, okay, these two are really strong. They came from really good companies. So find stripe. And okay, how do we keep that bar up? maybe and we didn't hire a lot waiting to find the right people and that paid off so you are slow at building the team which is trying to get the first 10 quickly yeah and we're just like working a lot ourselves yeah and how was your interview process

especially when you were trying to hire someone from Brazil or why or these different locations like how did you actually test for that quality bar so we actually brought them here Okay. So it's not, we, we didn't interview them remotely. So we had like technical interview. We had our, we're a little bit debatable, but like we believe that technical interview should be most similar to what the person will do in the day to day. And should test the knowledge that they would most have to do most decisions with. So we built like this interview that actually.

they had to bill like a small credit card authorized here, which like a part of it. And we could, it was the same interview for someone junior and someone senior. The only difference was how well they did the architecture and like the edge cases based on how junior they are and the senior they are. Like if they were senior, they knew like everything that would go wrong and they would build it or point it out if they didn't have time. And if they were junior, they would build a simple version of it. And we could tell based on that.

But then I think the hard part was the selling, right? Like how do you sell and get these people across the line? And I think that for the first 10, it was really hard. And we just like people joined who believed in the team and believed in the mission. And we never made compensation a problem. We were always very generous of the initial packages. So we were like, so we have this really different offer mechanism that, and this is until today. Everyone that joined a company went through.

Yeah. Can you talk about that a little bit? It is different and it's very intriguing. Yeah. We call it a pre-offer, not an offer, which we say, hey, if compensation wasn't an issue, you have all these companies you could join, would this be the company who would join? And let's discuss that. Until you tell me that this is the company you would join if compensatory was solved, like, I'm not going to show you an offer. And that made the person get over the mental decision of like, this is the company I want to join. And this is why it's better.

And then we always gave a super generous offer. It's not like we've been low-balled them, you know. We always gave them a great offer. And they were always like very few times we hadn't negotiated. Usually it was like above their expectations. And then they joined. But we've never had someone that after they accepted a pre-offer, they didn't come because of compensation.

We could always find a middle ground in compensation. I think there are a couple of other YC startups that actually use a similar approach, but the large majority doesn't. Yeah, it's definitely sometimes humbly Because people sometimes say no to you Before even seeing the offer Yeah But like I'm glad that that happened

Yeah Because they should go to the company That they believe more Right? And then we were evaluating And see what was their problem Maybe they weren't interested in the domain Maybe they didn't believe us as much

Like whatever it is We try to learn from it But we know that it was about the company Not about a comp You know like all those things Imagine if they had come Just before the comp Yeah

What problem would that be? Yeah. One of the other things I've noticed, especially working with you guys closely, is both you and Pedro do spend a lot of time on recruiting. And, you know, and first time founders have seen generally, you know, don't do that, especially in the early days and then wait too long before they actually refocus their attention on recruiting. Can you talk a little bit about why you guys do that? And you guys do it pretty well, which is like you spend a lot of time recruiting. And also talk about how much time you spend recruiting. Yeah.

So we spend probably 35 to 40% at our time recruiting. And that doesn't mean sourcing, by the way. Yeah. We think we really like our, well, we haven't hired a recruiter until now, but we had recruiters outsource working for us. I think since we were like four people. And we're big for, like we don't mind paying recruit fees if it's success based only. We don't like paying fees that like we have to pay before.

But like if it's success based, we think it's worth it. and we we we think that we read this in that google book i think that irish smit wrote that like don't let the urgency of the hire like reduce the bar of the people and we've done that so many times back in the day because like you always say hey well i'm just going to hire someone later i don't need that person now what this person like they're going to come and they're going to do what like there's all these arguments and the end of the day we just believe that to find the best people just have to interview more people like that's just what you have to do and if you really need them in like a month or two, you'll just won't interview that many people because

you'll just like, okay, I kind of need this person because that's going to make. And I think it's like a lot of times if you do need that person that month and it can make or break the business, you should hire them. But we just don't want to be in that position. So that's why we spend a ton of our time meeting a lot of people in recruiting and all of that. So can you talk about after you went through YC in Winter 17? What was sort of your fundraising history, both angels, your CDC and CDC B. Yeah. So we had a series A kind of together for seed round in March last year.

And that was I think $7.5 million round led by a fund called Ribit. That was after YC, right? And I remember we closed that round the morning of Demo Day. So it was very timely. And that was when Peter and Max came in and like a lot. lot of other investors. And then we we didn't fundraise. I have a little bit of different opinion fundraising than a lot of people.

I think you should build relationship of VCs over time instead of just doing a fundraising process and then like stopping. Because I think that if you're going to add someone to your life for like 10 years, 20 years, like you might as well know them outside of fundraising process. Yeah. Right. So we did actively meet a bunch of people and like started having even a week after finished fundraising.

We kept taking the meetings and like say, hey, we're not fundraised, but I would love to like Getting to know the partners, basically. I would love to get to know you. And like we met them three like every three months or so. And when the bee came around that we went with you guys, we already knew like everyone we wanted to talk to because we already knew people we like. and that treated us super well. And people that we didn't like that didn't treat us super well.

And then it was like a really fast process and because of that. Yeah. And I think one thing I would add there, this is where I think it's different for fintech and maybe certain select startups. Because, you know, even though you guys are not launching publicly until now, I think the key difference is, you know, you were working on building the tech stack, the fintech stack. And also you had at least more than 100.

pilot customers, right? So it was already being used by quite a few startups. And, you know, it's very similar to, if you look at the stripe of the Gusto story as well. You know, you can't, there's no room for errors in a payroll product. You know, it has, your paycheck has to come. And so, you know, Gusto sort of did the same thing, which is like almost for the first 15 months. They didn't publicly launch, but that doesn't mean they weren't testing the product. Exactly. Exactly. With customers, right. And so that's sort of what you guys were focusing during those 12, 15 months is what it takes to launch at scale by working with a few select startups.

Exactly. And I'm not saying we're not going to have bugs, but we spend a lot of time trying to reduce the possibility of having them. Because you're okay if your restaurant app doesn't work, right? But you're not okay if you're buying a dinner for a client and your car doesn't go true. Like how piss would you be, right? Yeah, that's true. And also like if the payments don't happen on time.

Exactly. Yeah. Yeah. So I think that's an important distinction for any. anybody who is working in fintech or a product where, you know, you don't have as much room for error. And so, you know, it comes down really to found a market fit and then product market fit and then seeing how to scale. And the other thing that's also true is for both Gusto, Stripe and Oz, which is different from a lot of companies as that, everyone needs a version of you. Everyone needs a version of payroll.

Everyone needs a version of payment processing. And everyone needs a version of corporate credit cards. you just has to be better than the other guys, which is very different than like, I don't know if I need this product that optimize my lead or whatever, you know? There's like, it's a little bit different concept when you're rebuilding something that already exists

in a better way that everybody has to get one or when you're creating a new market or when you're like just building something better of something that already exists. Yeah. It's different set of challenges and different set of it. Yeah. And the stack is pretty complex, right, if you want to have to build it from scratch.

Exactly. Exactly. So this is your second fintech startup, pretty much, right? And there are lots of founders trying to build something new in the fintech space, especially these days. I mean, we can see that even in the volume of applications that YC is getting. What advice do you have for someone who is a new founder trying to build a fintech startup? So being super transparent on it is I don't think we could have built Brex if you were, have not built a fintech startup before. Or at least worked at an earlier stage fintech started before.

Because a big part of it is that when you get to a like a meeting of a bank, like you know what you're talking about. Like you know how things work. You have credibility. We wouldn't be able to raise the size rounds we raised without like having that. So I would either work in a. a company in an early stage fintech company that you think it's successful or I would like yeah well I started a twin tech company that were but in Brazil is like a way less competitive market right like so if maybe you're international try to start something in your country

and then after move to the US I think in here in the US it's really really hard to do and I really admire people that had done it from the first time and and got it right like Patrick and John Yeah, like Patrick Gingon. But I don't think I personally could have done it without it. So I think like maybe joining would be a good idea. And for example, like people who join Pagarmé, like two of them started fintech companies are doing really well in Brazil, right, because they were able to get. And I'm pretty sure there's probably some Stripe alums that are starting fimp tech companies

that are going to do phenomenally well. What is difficult you think because of which working at a early stage fintech startup or things that you learn from Pagana? Like what are things that you think you will get core skills by working at a fintech early startup that you don't have when you're starting new? I think that there's a few things. One, understanding how financial systems work. So if you're an engineer, understanding how to build financial systems is different than building like a regular app or database. There's a lot of extra constraints, et cetera.

Second, there's just like knowledge about the market, how it works. And that you can learn. but it becomes really like the innovation within payments is a deep understanding of the constraints and a deep understanding of how to like go around those constraints. Can you give an example? So someone coming into the fintech market, for example, starting to build brex, what people would say would be, hey, what do you do is you go and you hire this like company called a processor. In this processor, they take care of like all these things for.

you and they're really good and you go to talk to the processor and they're really good they sell you that they do all these things like perfect like everyone does this seems to be a really good system just going to be an app on a mobile app on top of this and it's going to be great right and that's a common sense if you tell someone hey I'm going to build my own processing stack from scratch people will laugh at you they're going to be like no man you're crazy if you're asking anyone from any bang they're going to know that's impossible that's like super hard like you won't be able to do that and the thing is is that it is really hard but it's you know and if you don't do it you're going to be have a lot of constraints and if you haven't done it before

if you haven't seen the system that worked from scrap before and how that worked and why it's complicated it's actually is really hard to build one from scratch um so in terms of for example engineering is something like that or the other example is um so people tell you oh k yc which is like the um the way you have to get to your customer right for regulation purposes this is the way you have to do it you have to collect these documents etc etc but if you actually go read the law itself, it gives you a lot of flexibility in the way you collect information, the way you validate information. And if you're not aware of like, how can you go read that law, understand how that works, and apply it in that context, and then how do you sell the bank that that is a good thing,

like you won't be able to get away with it, right? It's not like you can create your inform and that's it. So I think just like being able to navigate all those things, it's not like you have infinite amount of shots. Like you have like one or two shots you have to get this thing right. So we have to have learned it from somewhere before. So it's a combination of knowing the regulatory requirements, understanding the complexity behind the financial systems and the tech stack,

and knowing that just plastering something on the top is not going to work. Exactly. And having the credibility for people to believe that you know that. Because you might know. But if nobody believes that you know because you've never done it, then it's going to be really hard. That's fair. That's fair.

And for you guys in the U.S., for sure, it's definitely harder coming from Brazil, but the fact that you build Pagahatatatmi helps you build that credibility. Because we actually knew all of our stuff, right? Yeah. And what does Brex exactly do today? Right? You obviously are just going to launch. So what is the value proposition for a startup founder on why they should use Brex over other options?

It sounds like getting a card is difficult. But beyond getting a card, like what are the things Brex actually does? Yeah, for sure. So we basically have two things going very quickly. The first one is you can get a card like from sign up to working virtual credit cards or credit card number to actually work and you can news online in like four minutes. Four minutes. Yeah. Okay.

So all the way from putting the entire information to getting a card like literally same day, four minutes. Exactly. Exactly. And so that was one. And today we're the only corporate card. It doesn't require any kind of personal guarantee or security deposit or anything like that. We underwrite 100% of the company, and we give you a limit that's many times higher than most of the banks would give. And you can get in four minutes.

So that's like one part of it. And the second part is that we automate a bunch of expense management stuff that you would have to do manually. And we just do all of that for you. And a bunch of expense management software can't do because they're not a credit card. I have to interact with the credit card. So we just build it all into the credit card. and you don't have to worry about a lot of things later. Later.

Let me touch on the first point because I think that is a big value proposition, especially to get the card pretty much in four minutes. And is that something the founders get, the entire company gets? Like, how does that work? Well, the founder has signed up gets the first one. And they can invite anyone in their company. And that's like the time literally to fill an username, password, and delivery address. And then they all instantly get a virtual credit card that works.

And they can set limits to them. So not everybody has access to all the corporate credit cards. You can give people a card of like $200 limits per person or whatever. Got it. And how are you able to underwrite do the KYC checks that quickly? Like what is different about the Brex stack that helps you do that? So this is one of the things that we feel very strongly about is when you're rebuilding, like when you want to disrupt, like, for example, I don't like disrupt. like a fancy word, but.

Redefining the experience. Yeah. When we're redefining the experience, like there's no way you can just build an app on top of an existing thing or building like a dashboard on top of an existing financial product. Like we believe you have to actually rebuild

the financial product. So that's like one of the things that we feel very strongly about. And when we started Brex, we're like, okay, we're not going to be like some sort of legion for an existing bank and they do everything. We're going to do everything from scratch. So in terms of, underwriting, we went and we rebuilt a whole

underwriting concept and instead of looking at financial history we look at cash, right? Like we look at how much investors actually invested in that company and went to write based on that and the average burn areas and all of that, which is something that it's specific for startups at this point, but it allows us to do it very quickly, right? So in terms of KYC, we just use modern methods to evaluate who you are and what

you do, right? Like, we use modern tools to not have to make you go to a branch and sign something physically. Like, we have better technology to have you sign something digitally. So, and the reason banks don't do that is because they all rely on these third party, like, old vendors that do everything for them, and they don't implement these new technology. So we basically just, like, rebuild the entire system and the entire stack from scratch, which allowed us to do these things. And is that the reason why you waited for launching? Because it's not, you know, most WIC startups launch after demo day, right? And you guys graduated and went to 17, but you're pretty much launching in mid-2018.

Like what was sort of holding you up from launching? It's exactly that. Like we didn't go to the strategy of like, hey, let's do something like and just launch it and have like whatever experience the bank gave us, you know? like we want to build something right and we had to take the time to basically rebuild all these systems from scratch right like we had to build a general ledger inside the company that controls all the balances so we don't lose money right those kind of things are really hard to build in the three months of yC and we believe that why a lot of fintech startups fail is because they don't take control over the full experience and they're always limited by the bank partner doesn't want to do this or the bank partner doesn't want to do that or this this old system doesn't want to do this. We're like, we're not going to deal with anything. We're just going to rebuild everything from scratch and

launch of that. Yeah. And one other thing I know we've talked about in the past, but I think it'll be helpful for, you know, our founder audience is like there's lots of lessons and learnings you have from paga.comi that you sort of brought in at Brex. Can you talk about especially setting up the process right? Because I know that's something. you felt very strongly about when you're launching Brex. And it's sort of how the product has been built to help the customer. Yeah. So the story on this is when we started programming, we didn't know this thing called

accounting existed. Like we didn't know it was a thing. Well, you were 15 and 16. Yeah. And they were like, oh, yeah. Okay, give me your P&L. I was like, what's a PNL? Oh, it's just like how much you're processing from his life.

I said, okay, I just build a spreadsheet. Hey, this is how much cash we had and just how much we burn, right? Yeah. You're like, no, I want a P&L. And you're like, okay. And then Pedro and I, we literally went to study accounting. And we literally did our books from scratch for three entire months because we had absolutely no accounting, right?

And that was like, oh, it was a great learning experience, but it took a lot of time and energy. And then when we got to Braggs, we're like, okay, we're not going to have this problem. We're going to have our accounting, like, really good from the first day. We're going to have, like, our expense management systems and, like, all these things, like, really well set up from the first day. Because if you know what you're doing, it takes a day to set up everything, right? It's like not that big of a deal. And it pays so much dividends later on.

Because like one thing that's really important is being able to control your business, like know how much you're spending. And that doesn't mean cash burn. Like losses and spend cash burn are very different things. Know how much that you're actually making, right? No on what you're spending is it. The categories, the vendors, like all of that. And that's a pain you has to do, right?

If you're going to do it manually with the existing. systems. Like I know it's really annoying and that's why like a lot of times founders just like don't do it and say, hey, when I raise my series A, when I raise my series B, I'll just pay someone to do it. But like it's not a type of thing also that you can throw money in a problem and then it's magically solved. Yeah. Right. Like you're going to have to spend your time on this and you just spend a day in the beginning setting these things up right. Like you won't have to deal with it in the future and it's like pay so much dividends. So I think that's one of the lessons. We just like build the financial part of the company, like, right from day one. Day one. So let's talk about the second element of your product, because the first element is the founder signs up. You're making it real easy for them to get access to card, as well as a certain credit limit because they have some cash in the bank.

They've raised from credible investors. So what exactly does Brex do to help set the process of accounting right for them? So basically a few things. The first one is we think it's important for a company to know how much they spend on a vendor, on a vendor base. How much you spend? When you say vendors like AWS, Google, AWS, Google, etc. Today, the status quo is you know how much everything you pay through ACH or wire,

you kind of know what you're spending on. But everything that you spend on card, you have no idea, right? Just credit card, $100,000 that have no idea what's in there. And then you have to consolidate all these extra reports from all these different services to know how much you're spending, right? We actually give you a report and upload to your accounting system, report on how much you're spending by vendor, which is not possible in the current cars today. The second thing is that, like, keeping track of receipts is something that if you don't enforce

the policy from day one, people just won't do it in the future. And you're going to have problems with that when you have to get an audit or, like, people are now have, like, all these expenses, et cetera. So we just created a super easy way. They can just, like, text the receipt to us, and that's it. We automatically match it to a transaction. And no, we don't have humans going manually doing it. It's actually automatic.

And the reason we could automate it is. versus other companies couldn't, is because we're both the credit card and we have the receipt. So matching your receipts with transaction is a way easier job than reading the receipt and figuring out everything about it. So that's like the second thing that we do

that it's super useful. And then we do all these things like we categorize everything, right? Because the automatic categorization, without getting too technical on payments, but like the automatic characterization from credit cards, they trust this thing called MCCs, and that's always wrong. Because so like, for example, Uber,

a lot of times comes just like car rental and you don't want Uber as in your car rental you want it as like taxi right so we redo that in the right way there's like a bunch of things we do that you just don't have to worry about it and we use it for ourselves and it's like we literally have no trouble accounting we

have our books every day it's like that's beautiful that's pretty awesome that sounds great and I know even though you are launching publicly only now you have been piloting your product for a while can you talk a little bit about your customers what types of customers use your product. Is this really early state startups or later stage? And how do the different use cases

work? So we have two use cases. We're usually on the earlier stage when we're people's first card. And they basically sign up for us or switch to us because we're super easy to get. We deal with all these accounting things that they don't want to deal with. And they don't want to deal with. And And you can just, like, don't not worry about it and, like, not have the hassle doing it. And we have, like, a lot of the R. YC batch uses it and the following YC batch, a bunch of people use it. So we got, we definitely use all the YC strategy. They're getting a lot of YC companies. And we're very, very happy about that.

And the other thing is, like, for companies that's over a certain size, we have another set of functionality that allows them to have better controls and, like, policies and stuff like that. So we got companies like firm or SOFi or color genomics to use Brex because they, when they get bigger, they start having problems with their current corporate card. So we help them add better controls and better reporting and a lot of larger company kind of functionality. And that allows us to scale with the company. So, yeah, we're pretty sure if like if someone starts with Brex, they can scale all the way with us. versus having to migrate to a different solution. Yeah. Yeah, one of the things I know, like, when we were working together on, like,

talking to some of the growth stage companies, their bigger pain points was like, with controls and especially as the employee-based expense, like, how do you know what's the spend on expenses versus accounts payables and so on? Exactly. Yeah. And so, you know, that was something that's still a pain point that's not solved. Exactly. So why do you think, I mean, why do you think this is getting solved now?

Why didn't it get solved? Was this not a problem 10 years ago? I think that B2B paying B2B with cards has grown a lot over the last five years. So if you think about like the largest company in the world, their main way to receive money is card. Right. Like Microsoft, Google, Facebook, you pay all your Facebook ads of card until you're like you're really, really big. And you could still pay off card, but it's a discussion for another point. And so I think that.

Before, like, card was just for T&E. Card was something that you paid, you gave your employees for travel, for paying cabs here and there. But now card is becoming this, like, procurement way of paying. So, like, way to pay, like, your actual big expenses. Like, a lot of offices are starting to take card. Like, we work, right?

Like, you can literally pay your card. So now it's getting to, it was always big, but now it's growing really, really fast and demanding new technology in the space to be able to scale to the company so it doesn't happen. Then, oh, now I agree. Now I have to move everything off of card and into invoices, right? So I think that changed over the last five years.

Yeah. And I also think with the new tech stack and the flexibility, like it's kind of absurd that you have a control or a set limit for the entire year versus is there something you can do more flexible at an employee level, right, for the company to better manage expenses. And especially startups where the cash bond really fluctuates. Exactly. Month to month.

Exactly. Well, thank you so much, Henry Kay, for joining us today. I'm sure that a lot of people probably really enjoyed the discussion, especially from how to start a fintech startup. Thank you very much. All right. Thanks for listening. So, as always, you can find the transcript and the video at blog.w.ycombinator.com.

And if you have a second, it would be awesome to give us a rating and review wherever you find your podcast. See you next time.