At nineteen I left Valladolid for San Francisco to build SDKs at startups I’d only met over email. A couple of years before that, still a teenager, I’d shipped a game called Infinity Field that reached the App Store Top 10 and got published by EA. I thought I’d already met the hard part of engineering. Fifteen years later I keep finding new ones, and chasing them turned out to be the whole career.
Since then I’ve worked on the parts of software that other software stands on. Compilers and language runtimes. Web frameworks that millions of developers ship on. A game engine. GPU rendering. Distributed infrastructure that handles billions of requests a day. The domain keeps changing and the appeal stays the same: I like the problems where the obvious approach falls over and you have to understand the machine to get back out.
Lately those problems are about AI agents. One of my favorites: I designed a batch system that consolidates an agent’s project memory the way sleep consolidates yours. Large offline passes run while the agent is idle, compressing and integrating everything it picked up during the day, so it wakes up knowing more and carrying less. Borrowing from how brains handle memory is the most fun I’ve had with a systems problem in a while.
What I’m working on now
I’m a principal engineer at Builder.io, where I set the technical direction for our AI agent platform across product, design, and code. Most of my time goes to three things: the developer tools and the experience around them, the applied-AI systems that make the agents genuinely useful, and the infrastructure that runs all of it.
That infrastructure spins up ephemeral Kubernetes clusters per workflow across AWS and GCP, tears them down when the work is done, and connects to enterprise networks through multi-tenant VPCs so customers can run agents inside their own walls. It serves billions of requests a day. The fun of it is the same as the fun of a compiler: getting a large, messy system to behave predictably under load.
Open source
Before Builder.io I spent years in open source, which is where work has to earn its keep in public.
- I created Gin, a Go web framework that became one of the most widely used in the ecosystem (around 88k stars and 290k+ dependent projects). More on how its router gets its speed in the Gin write-up.
- I helped take Qwik from proof-of-concept to production, including its Rust compiler and a resumability model that skips client-side hydration.
- I designed the compiler architecture and the high-performance virtual DOM for Stencil, and kept Ionic working across Angular, React, Vue, and plain JavaScript.
- Further back, I was a core contributor to the cocos2d game engine and shipped Infinity Field on top of it.
How I think about engineering
My CV from when I was a teenager has a line in it. “I want to make the best software ever, and build it so the people who use it love it.” I’ve grown more modest about the first half and more serious about the second.
The instinct that survived is treating performance and developer experience as features, not afterthoughts. Reducing draw calls in a game engine, tuning hot paths for V8 and JavaScriptCore, quantizing multi-gigabyte CT volumes so they render in real time in a browser, shaping an SDK so the first five minutes feel obvious. Same instinct, different layer.
Where to find me
I’m @manucorporat on GitHub and @manucorporat on Twitter, based between Lisbon and Budapest. If any of this overlaps with what you’re working on, email me at manu.mtza@gmail.com.