I build practical software where AI, product design, and systems engineering meet.
I like turning vague ideas into tools people can run, understand, and keep using.
focus Local-first AI apps, desktop tools, full-stack systems
stack Rust, Python, TypeScript, Tauri, FastAPI, Vue, React
style Product-minded, experiment-driven, maintainable
principle Make the hard part feel natural
- Local-first AI tools that keep user data on the user's machine
- Desktop apps with Rust/Tauri backends and polished web frontends
- Search, indexing, and automation workflows for real-world media
- Voice and vision experiments that become usable product flows
- Full-stack systems with clear boundaries, tests, and deployment paths
frontend Vue, React, Next.js, TypeScript, UI architecture
backend FastAPI, SQLAlchemy, Alembic, API design, job workflows
desktop Tauri, Rust, SQLite, background workers, local runtime setup
ai ONNX, GGUF, llama.cpp, FAISS, image/text/audio pipelines
quality ruff, mypy, pytest, vitest, eslint, typecheck-first habitsFirst external OSS contribution: a merged PyMuPDF pull request fixing a rendering metadata bug.
Public GitHub stats are only a small slice of my work.
I use this space for open-source activity and public-facing engineering traces.
building local AI tools that feel like normal software
learning better product shapes for voice, vision, and search
refining architecture that stays readable as features grow