- Henrique Dubugras co-runs Brex, the $12 billion corporate card and finance stack used by more than 30,000 startups.
- He founded his first fintech, Pagar.me, at 14 with best friend Pedro Franceschi and sold it for around $150 million in 2016.
- He dropped out of Stanford after eight months to join Y Combinator’s W17 batch and launch Brex.
- Brex survived the 2023 downturn, cut staff twice, exited the SMB segment, and returned to profitability with ~$500 million in annual revenue.
A $12 Billion Corporate Card Built for Founders Who Couldn’t Get One
In April 2026, Brex sits at a roughly $12 billion valuation, pulls in close to $500 million in annual revenue, and is profitable for the first time in its history. More than 30,000 venture-backed startups run their money through it — corporate cards with no personal guarantee, business banking, bill pay, travel, and expense management, all stitched into a single finance stack. DoorDash, Scale AI, and Anthropic were customers before they were household names.
The company is co-run by Henrique Dubugras and Pedro Franceschi, who share the CEO title and have shared most of their professional lives since they were teenagers on Brazilian internet forums. Dubugras is 29. He has been shipping code for twenty years, running companies for fourteen, and signing acquisition papers since he was legally a minor. The story of how a kid from São Paulo ended up running one of Silicon Valley’s most scrutinized fintechs starts with a desktop computer and a very impatient nine-year-old.
A São Paulo Kid Who Discovered Computers Before He Discovered Soccer
Dubugras was born in São Paulo in 1996 to two lawyers who expected him to follow them into law. He found a computer instead. At nine he was teaching himself HTML and CSS from Portuguese tutorials; by eleven he was building small websites for local businesses around his neighborhood and charging real money for them. He spent whatever hours were left playing video games obsessively, the kind of player who reverse-engineers game mechanics for fun.
He also got into trouble. As a pre-teen he was flagged for hacking into a school system he thought was poorly secured — a moment that pushed him toward building his own things rather than breaking other people’s. Around twelve, on a Brazilian forum for young developers, he met Pedro Franceschi, a Rio de Janeiro kid who had found a security flaw in Apple’s iMessage and been written up in TechCrunch. They started chatting daily. They have been building companies together ever since.
14 Years Old, a Best Friend, and a Payment Processor Called Pagar.me
In 2012, Dubugras and Franceschi founded Pagar.me, a Brazilian payment processor conceived as a local answer to Stripe. Dubugras was 14. Franceschi was 15. They wrote the backend between school assignments, registered the company under their parents’ names because they were too young to sign contracts, and began cold-emailing every merchant in Brazil who had a checkout form.
”Pedro and I built Pagar.me because we couldn’t find a single Brazilian company willing to process our own payments. We were 14. We had no other option.”
The two teenagers built the full payments stack themselves — acquirer relationships, fraud models, dashboards, APIs. By 2016, Pagar.me was processing enough volume to become the second-largest independent payment processor in Brazil, behind only the incumbent banks. They hired engineers twice their age. Dubugras was still finishing high school.
The $150 Million Exit at 18
In 2016, Stone, the Brazilian fintech giant then preparing for a Nasdaq IPO, acquired Pagar.me for roughly $150 million. Dubugras was 18. Franceschi was 19. They were suddenly rich, Portuguese-speaking, and out of school. Dubugras enrolled at Stanford’s computer science program on a scholarship that fall, partly because he wanted to understand how American startups worked from the inside.
He lasted eight months. The problem wasn’t the coursework. It was that he arrived in Palo Alto with real money, a real company idea, and no way to get a corporate credit card from any American bank. Every issuer asked for a US social security number and a personal guarantee tied to a US credit history he didn’t have.
”Stanford taught me what the American system looked like — and how badly it worked for anyone trying to build a company before they could legally drink.”
Eight Months at Stanford and One Unanswered Question
Dubugras and Franceschi dropped out together. Their first Stanford-era project was a VR startup called Beyond that went nowhere. Sitting in a Palo Alto apartment and staring at the same unanswered question — why can’t a funded startup get a corporate card the day it raises? — they pivoted. The new idea was underwriting on a company’s bank balance rather than the founder’s personal credit, cards shipped the same week a seed round closed, software built for engineers instead of CFOs.
They applied to Y Combinator and got in for the Winter 2017 batch. During the program they raised a seed round led by Peter Thiel’s Founders Fund, with Max Levchin and Ribbit Capital writing checks. Brex launched publicly in June 2018, and within six months nearly every YC company in the following batches was using it.
From $7 Billion to a Layoff Announcement to Profitability
Brex rode the 2020-2021 zero-rate boom straight up. A Series C in April 2021 valued the company at $7.4 billion. A Series D-2 in January 2022 pushed it to $12.3 billion, according to reporting from Forbes and Bloomberg. Then the rate environment flipped, startup deposits collapsed, and Brex’s SMB business began hemorrhaging money. In October 2022, the company cut 11% of staff. A second round followed in 2023, closer to 20% in the affected teams.
”We had to tell people we loved to leave. That’s the hardest thing I’ve done in business. I’ll never forget it.”
Dubugras killed the SMB product entirely and re-anchored Brex around venture-backed startups and growing enterprises — the customers who actually needed the full finance stack. By late 2025, the company was profitable. In early 2026 it reported roughly $500 million in annual revenue, with its valuation holding near the 2022 peak.
Building the AI-Native Finance Stack for the Next Decade
The fight now is with Ramp, and the weapon both sides are reaching for is AI. Brex has been rebuilding its product around an agent layer that automates expense categorization, policy enforcement, and receipt reconciliation in real time — the kind of tedious finance work that still ate hours at most of its 30,000 customers. Dubugras has been explicit that the roadmap assumes finance teams will shrink, not grow, and that Brex needs to be the system of record those smaller teams live inside.
For a co-CEO who shipped his first payments API at 14 and his first layoff memo at 26, the next chapter is less about building something new than about proving that a company born in a YC batch can stay sharp at scale. So far, the numbers say he’s getting there.
