Shipping cross-platform apps (Flutter, React/Web etc) with a strong product mindset and a soft spot for AI. I love turning real problems into useful, beautiful tools. I care about fast feedback loops, real user value, and shipping.
- building products for startups & smbs — fast MVPs → iterated v1s with clear metrics
- cross-platform UX — flutter for mobile, react/next.js for web; pragmatic stacks, clean APIs
- AI that earns trust — RAG, citations, eval harnesses, guardrails
I’m currently building internal software and data tools in the energy / grid space — think operator-friendly dashboards, data validation, and reporting workflows that support grid reliability and clean-energy integration. It’s a great playground for real-world constraints, scale, and safety.
tl;dr: i like hard problems with real users. and yes, I’m still open to building cool stuff on the side.
- open to freelance, collab, or small equity projects
- happy to advise on MVP scoping, product strategy, and tech choices
- especially interested in: AI assistants, ops tooling, health/med-device data, climate/energy software
Flutter · React/Next.js · TypeScript · Python/FastAPI · Firebase/Supabase · Postgres
LLMs (Gemini/OpenAI) · RAG · Vector DBs (FAISS/Pinecone) · Playwright/Scrapy · Docker
- Med-Device AI Q&A (RAG MVP) – Gemini-powered assistant over FDA 510(k)/PMA + internal docs.
Hybrid search, citations, safety rails. Cut research from hours → minutes. - CAISO-style Ops Reporting – modernizing legacy dashboards into a React front end with clean APIs.
Data validation, export pipelines, operator-friendly UX. - Lending Analytics – small suite of DS notebooks and apps for risk/return modeling.
Actionable visuals, human-readable explanations, reproducible code.
I value: clear problem statements, short cycles, and answers with receipts (links, citations, logs).
- 📊 Data Science for Lending → lutang123/Data-Science-Projects
- 📱 Flutter App Portfolio → lutang123/Flutter-MobileApp-Projects
- 🛠️ Full-stack Web Projects → lutang123/Web-Development-Projects
P.S. I’m packaging a slim RAG starter (ingest → index → retrieve → answer with citations). Ping me if you want early access.
Open to freelance / collab / full-time on: AI assistants, internal tools, data products, and “weekend-to-week-one” MVPs.
I’m endlessly curious, walkable-city obsessed, and I cook like I code: bold flavors, small batches, continuous improvement.


