I was 11 and I worked with my grandfather at his mechanic's shop. He had me take apart everything: engines, carburetors, gearboxes. One day I asked him: «Grandpa, why do you make me take apart everything?» He looked at me and said the line above. It took me twenty years to understand what he was teaching me.
Code came late. Not until 2020, at 23, did I pick up JavaScript on my own. No bootcamp, no university, no mentor — just the same shop logic applied to a different thing: take it apart until you get it, then put it back together better. Six years coding every day, two getting paid for it.
Today I'm still taking things apart, but now they're ideas. I take them apart to understand how to build them. A team that wants to add AI to their product but doesn't know where to start, a vertical no one attacked yet, an idea that exists in someone's head but not in production — they're all pieces you can imagine, design and build from scratch. The shop logic still holds: if you understand it, you can make it.
I worked at companies. They taught me to write tickets, wait for specs, hand off deploys. I learned the craft there. Then I found I move faster solo, closing the full loop. For the past 6 years I've been building products that way, end-to-end. LocalCenter, Sesión, JARVIS — they're not features I maintain, they're products that exist because I thought them, designed them, and shipped them.
Now I'm looking for a team where that matters. I don't want to be the engineer who gets tickets. I want to be the one who closes the loop: idea, code, deploy, customer, feedback, iteration. If your AI product needs to leave the lab and start charging money, I'm probably who you're looking for.