My Story

At 13 years old, I wrote my first real program — a Python script that renamed hundreds of files my dad had cluttering his desktop. It took an afternoon. That afternoon changed everything.

I didn't follow a tutorial or take a class. I Googled "how to rename files with code," copied a snippet, broke it, fixed it, and kept going. Within weeks I was automating anything that felt repetitive: scraping sports scores, generating homework templates, building a CLI calculator that nobody asked for.

What hooked me wasn't the code itself — it was the leverage. One person, one laptop, infinite output. I started seeing every manual process as something that should be automated.

I wanted to build things people could actually use, not just scripts that ran in a terminal. So I taught myself HTML, CSS, and JavaScript — the hard way, by building real projects with real deadlines.

A local restaurant needed a website. I said yes before I knew how to center a div. Shipped it in two weeks. It wasn't pretty, but it worked, and they paid me. That loop — say yes, figure it out, ship it — became my operating system.

By the end of the year I'd built sites for three small businesses, learned responsive design by breaking layouts on my phone, and discovered that client work teaches you more about software than any course ever could.

Frontend started feeling like decoration. I wanted to understand what happened after someone clicked the button — where data went, how it got stored, why things broke at 2 AM.

I went deep on Node.js, PostgreSQL, and server architecture. Built my first REST API that other people actually depended on. The moment I watched real traffic hit an endpoint I'd written was a different kind of satisfaction than anything frontend ever gave me.

The biggest lesson this year wasn't technical — it was learning that infrastructure decisions compound. A shortcut in your database schema today becomes a migration nightmare in six months. I started thinking in systems, not features.

I became obsessed with a question: how do you get multiple computers to agree on anything? That question led me down the rabbit hole of distributed consensus, and I haven't fully come back up.

Read the Raft paper, then Paxos, then the Git internals documentation cover to cover. Built toy implementations of each to make sure I actually understood them — not just the happy path, but the edge cases where everything falls apart.

This year reframed how I think about software. Every system is distributed if you zoom out enough. Networks fail, clocks drift, messages get lost. The interesting problems aren't "how do you build it" — they're "how do you build it so it stays built when things go wrong."

At 16, I incorporated my first company. I thought the hard part was building the product. I was wrong — the hard part was everything else.

Learned about product-market fit by not having it. Built features nobody asked for. Priced things wrong. Spent weeks on a landing page when I should have been talking to users. Every first-time founder mistake in the book, I made it.

The company didn't survive, but I did. And I walked away with something more valuable than revenue: a visceral understanding of what not to do. The next time I started something, I started by talking to people first.

October 2025. One company, four products, same driving instinct: the tools developers use every day are broken, and nobody is fixing them properly.

LegionEdge is the parent company. Every product underneath it tackles a different piece of the developer experience that nobody else is getting right.

W0rktree replaces Git entirely — a version control system written from scratch in Rust for monorepos, multi-tenant organizations, and real-time collaboration. Open sourced from day one.

Nokuva is the premier AI-native design editor. Design tools that jump straight from prompt to code skip the entire design phase. Nokuva puts design back at the center with a real canvas, a design token system, and multi-agent AI.

Foltrac tackles multi-cloud deployment. Every team I'd worked with had the same complaint — deploying across AWS, GCP, and Azure felt like maintaining three separate infrastructure stacks. Foltrac unifies that into a single orchestration layer.

Tavoc is the IDE I wished existed when working on large codebases. Purpose-built for monorepo navigation, cross-project refactoring, and sub-millisecond file switching.

Founding a company with four products at 19 meant learning to delegate, prioritize ruthlessly, and accept that "good enough shipped" always beats "perfect in progress."