I don't stop at a notebook.
I wire models into APIs and UIs, put them behind CI, and keep the deploy path boring on purpose — so the product still works on a Tuesday.
- ✨ AI in the product, not beside it
- 🛠️ Ship muscle — Actions, Docker, cloud when it earns its keep
- 📐 Small surfaces, sharp edges — fewer moving parts, clearer contracts
| layer | tools I reach for |
|---|---|
| product UI | TypeScript, React, Next.js |
| AI / API | Python, FastAPI, model APIs |
| data | Postgres, Redis when needed |
| ship | Docker, GitHub Actions, cloud |
swift-supabase-auth-http— Supabase Auth HTTP client (Swift)swift-coverage-action— coverage GitHub Action
(swap these for flagship AI full-stack repos as they land)
currently · training the full loop on this profile — AI feature → end-to-end → CI green
quiet craft · loud uptime



