AI Project Scoping Canvas
An infinite canvas where branching question nodes turn a one-line build idea into a structured map of assumptions, risks, and follow-up requirements.
Fuzzy workflows. Explicit systems.
Product engineer with a bias for typed contracts, useful tests, and dependable production paths.
I design the product surface and the systems beneath it: interaction flows, typed UI state, AI loops, and change-safe architecture that keeps teams shipping.
A lot of database code starts by selecting too much.
Before AI coding agents could read error logs, run commands, or inspect a codebase on their own, I was building a browser-based workflow around GPT-4 that tried to make the same loop faster.
React makes local state easy to start with. That is one of its strengths.
An infinite canvas where branching question nodes turn a one-line build idea into a structured map of assumptions, risks, and follow-up requirements.
TypeScript diagnostics CLI that turns verbose compiler output into concise semantic diagnostics for AI coding agents.
A TypeScript library that models complex UI behavior as an explicit state machine - every component binds to a tagged contract of data, actions, and reachable states.