The friendship that built Brex started with an argument on Twitter.
In 2012, Henrique Dubugras was a high school senior in São Paulo. Pedro Franceschi was a high school senior in Rio de Janeiro. Both were obsessed with code. Both had already done things that would have gotten them in serious trouble with adults paying closer attention. Pedro had been the first person in Brazil to jailbreak the iPhone 3G at 13 years old, which generated a patent infringement notice from Apple. Henrique had built a clone of a video game called Ragnarok in middle school, generated tens of thousands of dollars from it, and eventually received a legal warning from the game’s Brazilian distributor.
Neither of them stopped.
They found each other on Twitter when they got into a disagreement about the nuances of coding tools. The 140-character limit wasn’t enough space to actually resolve anything, so they switched to Skype. Within an hour they had become friends. Within a year they had started a company together.
The company was Pagar.me, built when Henrique was 15 and Pedro was 14. It was, in the most direct description, the Stripe of Brazil: a payments infrastructure company that made it easy for Brazilian businesses to accept online payments through a clean developer API rather than the legacy payment systems that dominated the market. They had seen how Stripe had changed payments in the US and believed the same infrastructure gap existed in Brazil. They were right.
By the time they sold Pagar.me to Stone, a Brazilian fintech company, in 2016, the company had 150 employees and was processing over $1.5 billion in transactions annually. The founders were 19 and 18 years old. The sale price was described as “tens of millions of dollars.”
They enrolled at Stanford that fall. They lasted eight months.
The VR Company That Wasn’t and the Problem That Was
Henrique and Pedro entered Y Combinator’s Winter 2017 batch with an idea for a VR company called Beyond. The logic was reasonable on its face: VR was the emerging frontier, there was genuine technical excitement around it, and they wanted to build something ambitious.
Three weeks in, Henrique went to the Consumer Electronics Show in January 2017 and came back with a clear read on the situation. He and Pedro didn’t know anything about the VR market. They were not the right founders for this company. They had come to the US assuming the payment infrastructure problems they had solved in Brazil were already solved here, so they had tried to find the next frontier. But when they looked at what was actually broken in their immediate environment, the frontier was right in front of them.
At YC, Michael Siebel asked them a pointed question that restructured their thinking: if you could build anything, what would you build? Their answer was a business bank for US startups.
The problem they were describing was something they had encountered personally. As YC founders with a funded startup, getting a corporate credit card was still an ordeal. Traditional banks evaluated credit applications using personal credit history, existing business revenue, or required personal guarantees that would make the founder personally liable for business debt. None of these criteria were relevant to the actual financial health of a funded startup. A company that had just raised $2 million from top-tier VCs had excellent credit quality by any reasonable measure, but that quality was invisible to a bank’s underwriting model because it showed up in bank account balance and investor backing rather than in credit history or revenue history.
The banks were not being irrational. They were applying models built for a different kind of business to a kind of business those models couldn’t read. Henrique and Pedro understood how to build the model that could read it because they had built payment infrastructure from scratch before and understood what data was actually predictive of credit quality for early-stage companies.
They scrapped Beyond in April 2017. Brex was founded that month.
Building the Card From Scratch
Pedro Franceschi built Brex’s backend from scratch. The core card processor. The KYC (know your customer) functionality. The underwriting engine. The connective tissue to Visa and Mastercard.
This was a deliberate choice and an unusual one. Most fintech startups building card products in 2017 used third-party processors and existing banking infrastructure as their technical foundation, accepting the constraints and limitations of those legacy systems in exchange for speed to market. Pagar.me had taught them what happened when you built on other people’s infrastructure: you couldn’t modify the important things. Underwriting parameters, payment terms, rewards structures, the things that would differentiate Brex from every existing corporate card, required owning the underlying technology.
Building from scratch took longer. It produced something that Brex could actually control. When they later needed to make credit decisions in real time using live bank account data rather than credit history, or when they needed to set credit limits based on a company’s funding and cash balance rather than revenue history, those capabilities required the kind of architectural flexibility that third-party infrastructure couldn’t provide.
The underwriting model they built was the core differentiator. Traditional corporate cards required personal guarantees, meaning the founder put their personal credit on the line for the company’s spending. Brex required no personal guarantee. Instead, they evaluated the company’s financial health directly: how much money was in the bank, where that money came from, what the company’s trajectory looked like based on the data they could see. A startup with $2 million in the bank from a top-tier VC fund was a better credit risk than the personal credit score of its 24-year-old founder would suggest, and Brex’s underwriting recognized that.
They could also offer limits dramatically higher than traditional options. Where a startup founder might get a $5,000 to $10,000 limit on a traditional corporate card or business credit card, Brex could extend limits as much as ten times higher based on the company’s actual financial position. For a startup trying to run a real business, the difference between a $10,000 credit limit and a $100,000 limit was the difference between constant manual expense management and being able to actually operate.
The application experience was another design decision made with the target customer in mind. A founder could apply, be approved, and have virtual cards issued within minutes rather than the weeks that traditional bank applications required. Automated receipt capture reduced the manual burden of expense tracking. A consolidated view of company spending, organized by vendor and category, let a CEO understand at the end of each month how much the entire company spent on Uber or AWS without manually reconstructing it from individual receipts.
YC, Billboards, and the Demo Day Distribution Network
Brex’s first significant growth channel was the YC alumni network, and it was self-generating rather than engineered.
The product solved a problem that every YC-backed startup had encountered. Every YC founder knew that getting a corporate card was harder than it should be. When Brex launched, the YC community was the most concentrated population of people with the exact problem Brex solved, and they were tightly networked enough that word traveled fast. Within the YC ecosystem, Brex spread through referrals without requiring a formal sales motion.
Weeks before Demo Day in 2017, Ribbit Capital led a $7 million Series A into the company before it had even publicly launched. Micky Malka at Ribbit would become Brex’s biggest shareholder and board member from the outset. The Series A before Demo Day was an unusual signal of investor conviction, driven by both the founders’ track record with Pagar.me and the clarity of the problem Brex was solving.
In June 2018, Brex launched publicly. The growth strategy combined several channels: outbound sales targeting foreign founders in the US who faced particular difficulty getting credit cards without US credit history, billboard advertising in San Francisco spending approximately $300,000 on outdoor placements across the city, referral programs, and brand awareness campaigns designed to make Brex synonymous with “the corporate card for startups.”
The billboard campaign was both a literal marketing investment and a brand statement. Brex cards appeared across San Francisco in a city that was saturated with fintech advertising, but the specific positioning, a corporate card for startups that understood how startups worked, cut through in a way that category-generic financial advertising didn’t. Founders who were dealing with the frustration of trying to get banking services that matched their company’s actual situation saw Brex on a billboard and understood immediately what the pitch was.
The 2018 Series C of $150 million from Kleiner Perkins at a $1.1 billion valuation made Brex a unicorn less than two years after founding. The founders were 22 years old. The narrative was clean and viral: two Brazilian teenagers who had met on Twitter, built and sold the Stripe of Brazil as minors, arrived in Silicon Valley and became billionaires before they could legally rent a car.
From Startup Card to Financial OS
The corporate card was the wedge. What Henrique and Pedro were building toward was the full financial operating system for a startup.
The insight was structural: every startup needed financial infrastructure from day one. Bank account, corporate card, expense management, bill payment, cash management. These were not optional additions to a startup’s toolkit. They were the foundation. If Brex could get into a company at the moment of founding, through the corporate card that every funded startup needed immediately, it could expand from that initial relationship to own more of the company’s financial workflow over time.
The product expansion followed this logic. Brex business accounts offered startups an alternative to traditional business bank accounts, with higher interest rates on deposits, no minimum balance requirements, and integration with the card product that eliminated the friction of reconciling between separate institutions. Expense management software turned the raw transaction data from the card into an organized view of company spending, with automated categorization, receipt matching, and policy enforcement. Bill payment automated the process of paying vendors. Financial modeling tools helped finance teams with forecasting and budget management.
By 2021, Brex had launched venture debt, offering startups additional non-dilutive capital. By 2022, Brex Empower packaged these capabilities into a financial software platform specifically designed for enterprises managing complex, multi-department spending.
The framing that Henrique used was precise: Brex had a very similar effect on startup finance as what Stripe had had on payments in the beginning, but moving much faster because Silicon Valley companies were very good at spending money. The interchange revenue model meant that Brex grew with its customers’ growth: the more a startup scaled its operations, the more it spent on its Brex cards, the more interchange revenue Brex captured.
By 2021, 80% of YC companies used Brex. One-third of the portfolios of the top 50 venture firms were on the platform. A quarter of all US startups had Brex cards.
The $12 Billion Valuation and What Came After
January 2022. Brex’s Series D-2 round, led by TCV and Greenoaks Capital, valued the company at $12.3 billion. Total funding raised had reached approximately $1.5 billion across all rounds.
The valuation was a product of a specific moment: zero interest rates, abundant venture capital, startups raising at historically high valuations and spending freely, and a financial technology sector that was being valued on revenue multiples that assumed the growth rates of 2020 and 2021 would continue indefinitely.
Several things changed in 2022.
Interest rates rose. The venture funding environment contracted sharply. Startups that had raised at high valuations began cutting headcount and reducing spending. The specific relationship between Brex’s revenue and startup spending meant that when startups pulled back, Brex felt it directly: companies were issuing fewer cards, making fewer purchases, generating less interchange revenue.
The harder problem was structural. Brex had expanded to serve small and medium businesses beyond the venture-backed startup ecosystem, and those customers had different needs, different unit economics, and different support requirements. In June 2022, Brex made the decision to exit the SMB market, informing tens of thousands of small business customers that their accounts would be closed unless they met new eligibility criteria requiring VC backing or over $1 million in revenue.
The execution of the announcement was widely criticized. Customers who had built their businesses on Brex infrastructure were informed by email with relatively short notice. Social media erupted. The founders published a public mea culpa. Internally, employees described the period as chaotic and the strategic direction as uncertain. Eleven percent of the workforce was laid off later in 2022.
The strategic logic behind the SMB exit was defensible even if the execution was poor. Brex’s revenue fell less than 2% after letting go of those customers, which confirmed they were not meaningfully contributing to the business while consuming significant support resources. The company’s product and underwriting model was built around companies with institutional backing, and trying to serve the broader SMB market with that product created friction in both directions.
In January 2024, Brex laid off approximately 20% of its workforce, roughly 300 people, citing slowing growth. The same announcement included the statement from co-CEO Pedro Franceschi that Brex had grown gross profit by 75% the prior year, pointing toward a company that was getting more efficient even as it was getting leaner.
The SVB Moment and the Repricing
On March 9, 2023, the day before Silicon Valley Bank collapsed, Brex received billions of dollars in deposits from SVB customers who were moving their money to safety. The moment was a validation of what Brex had built: when the most prominent bank serving the startup ecosystem failed, founders chose Brex as the alternative. The trust that had been built through years of serving the startup community translated directly into deposit inflows at the moment it mattered most.
The SVB collapse accelerated Brex’s business account growth in a way that no marketing campaign could have produced. Thousands of startups who had relied on SVB for their banking needs needed an alternative immediately. Brex was the obvious choice for companies that had already been using Brex cards and knew the product.
In early 2024, Brex undertook an unusual internal restructuring of employee equity. The company repriced equity to approximately $4 billion, well below the $12.3 billion peak valuation but also well below what the market was pricing Brex’s equity at in secondary markets. The repricing was unpopular with some investors and acknowledged the gap between the 2022 peak valuation and the realistic business trajectory. Pedro Franceschi’s argument was direct: employees needed to fundamentally believe in the value of their equity. Maintaining compensation tied to a $12 billion valuation that the business’s fundamentals no longer supported was not serving anyone’s interests.
By 2025, Brex had rebuilt its revenue trajectory. Annualized revenue reached approximately $700 million, up from a low of $312 million in 2022. The company was approaching cash flow breakeven. The enterprise pivot, despite its rocky execution, had moved Brex toward customers with more durable revenue characteristics than early-stage startups: companies like DoorDash, Coinbase, Scale AI, Anthropic, Robinhood, and Zoom had become major customers.
Capital One, $5.15 Billion, and How to Think About the Exit
In January 2026, Capital One announced the acquisition of Brex for $5.15 billion in a 50/50 mix of cash and Capital One stock. The deal is expected to close in mid-2026, subject to regulatory approvals.
The story has two framings and both are accurate.
The glass half empty version: $5.15 billion is a 58% discount from the $12.3 billion peak valuation. Investors who came in at the 2022 round paid prices that implied outcomes well above $5 billion. Kleiner Perkins, which led a $100 million round at a $2.5 billion valuation in 2019, will make roughly 1.2 to 1.3 times their money after dilution, a return that is below what a successful growth equity investment should produce over six years. The rival outcome, Ramp raising at a $32 billion valuation in November 2025 and announcing over $1 billion in annualized recurring revenue, makes the comparison sharper.
The glass half full version: Ribbit Capital’s Micky Malka, who led Brex’s $7 million Series A in 2017 and became the company’s biggest shareholder, is looking at a return that is unambiguously spectacular. Early YC investors made similarly strong returns. $5.15 billion is not sympathy money: it is half cash and half liquid public stock in Capital One, one of the largest banks in the United States. The founders built something from a Twitter argument in 2012 to a $5 billion exit in nine years, starting as teenagers with no institutional backing and doing it twice.
Pedro Franceschi’s framing in his post-announcement interview was straightforward: Brex would operate as an autonomous unit inside Capital One, with Pedro remaining CEO, with the backing of a $130 billion bank’s balance sheet, $6 billion in annual R&D spending, and the largest credit card distribution network in the US. The financial constraints of operating as an independent company in a capital-intensive business would be removed. The opportunity to reach scale that no standalone fintech had achieved, competing directly with American Express and JPMorgan Chase for corporate card dominance, would be accelerated.
Capital One’s CEO Richard Fairbank had just completed the $35 billion acquisition of Discover Financial, giving Capital One access to one of the only payment networks at any scale. The Brex acquisition added the technology layer: a cloud-native, AI-powered expense management platform built from scratch with the kind of engineering quality that Capital One’s own technology modernization had taken years to approximate.
Whether the combined entity makes the Capital One-Brex deal look prescient in five years depends on execution. But the terms of the deal, 13.4 times forward gross profit at the top end of public fintech comparables, are not the terms of a distressed sale. Brex had rebuilt its fundamentals sufficiently that it could negotiate from a position of business strength rather than existential pressure.
What the Brex Story Is Really About
The Brex success story contains several different stories depending on when you start and stop reading it.
If you read it from 2012 to 2018, it is a story of two Brazilian teenagers who met on Twitter, built the Stripe of Brazil as minors, sold it, arrived in Silicon Valley without knowing what they would build, watched their YC batchmates struggle to get corporate cards, and built the infrastructure to fix it in under two years.
If you read it from 2019 to 2022, it is a story of hypergrowth and the specific risks of building a business whose revenue is structurally correlated with startup spending at a moment when startup spending was approaching its all-time peak.
If you read it from 2022 to 2024, it is a story of a hard pivot executed imperfectly, a valuation correction that was painful for late investors, and a company rebuilding its fundamentals under real financial and organizational pressure.
If you read it from 2024 to 2026, it is a story of a company that rebuilt to $700 million in revenue, achieved cash flow breakeven, navigated the SVB crisis as a beneficiary, and found an acquirer willing to pay $5.15 billion to acquire what it had built.
All of these stories are true simultaneously. The Brex story is useful specifically because it contains the full cycle: founding insight, hypergrowth, valuation inflation, correction, pivot, rebuild, and exit. Most startup stories are told from one moment in that cycle. Brex’s story got to finish.
Pedro Franceschi was 13 when he jailbroke the iPhone 3G in Brazil. He was 14 when he and Henrique started Pagar.me. He was 19 when they sold it. He was 21 when they pivoted from VR to corporate cards inside a YC batch. He was 22 when Brex became a unicorn. He was 28 when the $12 billion peak valuation. He was 30 when he signed the term sheet with Capital One on December 22, 2025, and announced the $5.15 billion deal a month later.
From a Skype call in 2012 between two high school students in São Paulo and Rio de Janeiro, arguing about code, to the largest bank-fintech acquisition in history.
That is what the Brex success story is really about.
Leave a Reply