i met guilherme lima at klivo. he was a product manager, i was an engineer. klivo helped people with chronic conditions live better — think livongo, but for brazil. we spent two years there, and the most valuable thing about those two years wasn't the work. it was the conversations. we'd talk for hours after work about thesis ideas, healthcare, the vc game, what makes a product good, what makes it last. gui is one of the smartest people i know. by the end of those two years, i knew that if i was ever going to build something, it would be with him.

but we weren't ready yet. i was 22 and tired of healthtech. i wanted to go deep on the technical side, write code without thinking about patient data. so i left klivo and joined cloudwalk, a fintech in brazil, as a staff engineer. gui left too. we both needed a break from healthcare.

that lasted about three months.

gui called. "i need you to leave cloudwalk. i'll leave my company. we'll build something together. i don't care what it is."

so i quit. left a well-paying job with no funding, no product, no plan. just a cofounder i trusted and a feeling that the window was now or never.

we spent weeks exploring different ideas. one kept coming back: building infrastructure for healthtechs in brazil. something like what truepill was doing in the us. but as we dug deeper, we noticed something that reframed everything.

brazil's open banking ecosystem was already mature. banks were sharing data, apis were standardized, fintech was thriving. but healthcare? completely disconnected. patient data scattered across hospitals, pharmacies, labs — none of it talking to each other. you'd do a blood test at one lab, visit a specialist at another hospital, and get a prescription at a third pharmacy. none of them knew about the others. your medical history existed in fragments across dozens of systems that would never speak to each other.

the question felt obvious once we saw it: why not build open health?

a health intelligence platform that connects electronic health records, hospital systems, labs, and pharmacies into a single network. patients consent to share their data once, and from that point their full medical history is accessible wherever they go. one identity across the entire health chain. no more filling out the same forms at every new clinic. no more repeating blood tests because the results live in some other system's database.

for patients, it meant having something like an oauth for healthcare. log in once, carry your records everywhere. for insurers and health plans, it meant they could finally underwrite and price plans based on real, complete health data instead of fragmented paperwork and self-reported forms. for providers, less duplication, better outcomes, faster decisions. the incentives were aligned for everyone in the chain — which almost never happens in healthcare.

we built it in a couple of months. raised a pre-seed round from angels. started getting our first customers. on paper, everything was working.

in practice, we were exhausted.

the problem wasn't the product. the problem was the sales cycle. enterprise healthcare in brazil moves slowly. we had companies interested, but "interested" in this market means six months to a year before a contract closes. we were just waiting — building, improving, waiting — for decisions that were always "coming next quarter."

i was on a call with the ceo of axenya — someone we knew from the industry — and i said something honest about our frustration. we had amazing tech. we had a real thesis. but we had maybe a handful of companies in brazil that would actually buy it on a timeline that kept us alive.

he offered to buy us.

we didn't say yes immediately. gui and i spent weeks going back and forth. we'd built this thing from nothing — quit our jobs for it, stayed up until 4am debugging integrations with hospital systems that ran on software older than we were. saying yes meant letting go of the version of the future where healthapi becomes the plaid of healthcare in brazil. that's not easy to walk away from, even when the rational answer is obvious.

but the more we talked about it, the more we kept arriving at the same place. axenya already had the distribution. they had the sales team, the enterprise relationships, the regulatory credibility that takes years to build. our tech inside their machine would reach more patients in six months than we could reach on our own in three years. and honestly — we were two guys in their early twenties trying to sell to the most conservative buyers in the country. the tech was never the bottleneck. access was.

we said yes.

the whole thing — from gui's phone call to the acquisition — was about eighteen months. it was the most compressed education of my life. we saw a real problem, built real technology to solve it, got it into the hands of real customers, and found the right home for it. not every company needs to be a decade-long journey. some of the best ones are fast.

thank you gui, for all the crazy days we spent together.