Szymon Rybczak was shipping code to React Native at 16 and maintaining libraries millions of developers rely on. Then he got into Y Combinator - while still in high school - and moved to San Francisco to build TesterArmy.
Most people spend their teenage years waiting for permission - to drive, to vote, to be taken seriously. Szymon Rybczak spent his shipping code that millions of applications quietly depend on. By the time he was old enough to be legally interesting, he had already contributed to React Native, maintained open-source tools with millions of weekly downloads, and talked his way into Y Combinator without a high-school diploma in hand.
The story reads like it should have several more decades attached to it. It doesn't. And that gap - between what he was allowed to do and what he actually did - is the most interesting thing about him.
Rybczak started programming around the age of 14. That part is common enough; plenty of kids write their first line of code at that age. What is uncommon is what he did next. Within two years he was contributing to React Native, the framework Meta built to let developers write mobile apps once and ship them everywhere. He became one of its youngest-ever contributors at 16.
To understand why that matters, you have to understand what open-source contribution actually is. It is not a class project. It is submitting work into a system where strangers - senior engineers, maintainers, skeptics - review every line and can reject it without ceremony. A teenager doing this successfully, repeatedly, is not a novelty. It is a signal.
He took that signal to Callstack, a well-known React Native consultancy in Poland, where he became the youngest developer on the team. Over roughly three years there he worked on the kind of infrastructure most engineers never touch: the React Native Community CLI and react-native-ai, libraries with more than two million weekly downloads between them. When your code sits under that many other projects, a small mistake ripples outward fast. He learned to work carefully, in public, under review.
There is a version of the founder story that skips the boring middle. Rybczak's middle was the whole point. Before he pitched anything, he spent years giving conference talks - explaining React Server Components to a room of senior engineers at React Day Berlin in 2023, walking audiences through on-device language models at React Advanced London in 2025. He was, in effect, teaching the ecosystem while learning its hardest corners.
Companies Rybczak has built alongside co-founder Oskar Kwaśniewski - a partnership that started years before either could rent a car.
That last detail is easy to skim past. TesterArmy's three founders - Rybczak, Oskar Kwaśniewski, and Piotr Matyjasik - are all from Poland. Rybczak and Kwaśniewski had worked together across four different companies before this one. Piotr and Oskar studied computer science together. When they finally started a company of their own, they were not strangers gambling on a pitch deck. They were people who already knew how each other worked under pressure.
The idea behind TesterArmy is almost annoyingly simple to state, which is usually a sign it took real work to reach. Every engineering team knows they should test their software more. Almost none do enough, because writing and maintaining automated tests is tedious and it breaks constantly. So the founders built an AI agent that behaves like a real user: you describe what to test in plain English, and the agent clicks buttons, fills forms, and navigates your app using vision. No brittle code selectors. No scripts to babysit.
When something goes wrong - a broken flow, a rendering bug, confusing copy - the agent hands back screenshots, recordings, and a plain report of what it found. It plugs into GitHub and Vercel with one click and runs on every pull request. By the time the company launched publicly through Y Combinator, it was already running in more than 20 production companies.
Getting into Y Combinator is hard for anyone. Getting in before you have finished high school is a different category of thing. When the acceptance came, Rybczak posted it plainly on LinkedIn - still in high school, building a devtool company, moving to San Francisco. The post drew nearly 700 reactions and a wave of congratulations from people who had watched him grow up in public, one commit at a time.
The decision underneath the post is the part worth sitting with. He could have finished school the conventional way and kept building on the side. Instead he and his co-founders packed up and relocated a company and their lives across the world. Some choices are agonizing. This one seems to have been easy, in the way things are easy when you already know exactly what you want to do.
It would be easy to file Rybczak under prodigy and move on. That misses the mechanism. He is not interesting because he is young; he is interesting because he spent years doing unglamorous work where anyone could watch - fixing bugs, maintaining tooling, explaining hard concepts on stage - until the compounding did the loud part for him. Open source was his resume. Y Combinator simply read it.
His stated ambition is straightforward and large at the same time: change how AI-native teams test their software, and make TesterArmy the default layer that catches problems before customers do. Whether he gets there is an open question. What is already settled is the pattern - pick a hard problem, stay with it in public for years, and let the work introduce you. He has done it once. The company is the second run of the same experiment.