Skip to content
Dylan Taylor

About

Dylan Taylor

I'm 17, based in Oakland, CA, and I build AI tools — mostly focused on making their answers reliable and easy to check. I design and ship real projects using AI to help me write the code, I understand how backends and system architecture fit together, and I can read and follow almost any codebase.

I'm strongest at prompting, product design, and getting a rough idea into something that actually works. I'm still leveling up my from-scratch coding. I've been teaching myself since I was nine, and I don't quit until a problem is solved.

What I work on

Three things at the moment. Littman is AI for legal work that returns cited answers and routes the uncertain ones to a person. OENRA reads a wine's chemistry and a person's palate on the same eight dimensions and explains the match instead of scoring the bottle. FlightReady scores a pre-flight go/no-go decision by its tail risk rather than its average.

The through-line is decisions somebody has to defend afterwards — to a partner, a regulator, or themselves. That constraint changes what you build. It makes citation, uncertainty, and the ability to abstain part of the product rather than polish applied at the end.

Before this

From about nine to thirteen I taught myself C and C++ by reverse-engineering game binaries, working against kernel-level anti-cheat teams. I stopped on my own once I realized I'd rather build things for people than against them. The habit of pulling hard systems apart is still how I approach AI.

Elsewhere