Your 100% local AI Second Brain that actually remembers everything. Draft notes in markdown → save to your private vault → chat with an AI that pulls perfect context from your own thoughts. Zero cloud. Zero subscriptions. Zero data leaks.
- 100% Local & Private — Everything runs on your machine with Ollama
- Capture Zone — Dual-pane markdown editor with instant save
- Smart RAG Chat — ChromaDB finds relevant notes automatically
- Real-time Streaming — Beautiful SSE streaming responses
- Persistent Everything — Notes + chat history survive reloads
- Auto-Healing Vectors — Edit a note → old embeddings are intelligently replaced
- Life OS Ready — Built for daily thinking, not just note-taking
- Frontend: Next.js 14 (App Router) + Tailwind + shadcn/ui
- Backend: FastAPI + LangChain + ChromaDB
- AI Engine: Ollama (
llama3.2+nomic-embed-text)
- Node.js (v18+)
- Python 3.9+
- Ollama (running)
ollama pull llama3.2
ollama pull nomic-embed-textcd backend
python -m venv venv
source venv/bin/activate # Windows: venv\\Scripts\\activate
pip install -r requirements.txt
uvicorn main:app --reload --port 8001 --host 127.0.0.1cd frontend
npm install
npm run devOpen http://localhost:3000 → start building your Life OS!
/vault← Your actual .md files + hidden .chroma DB (auto-created)/backend← FastAPI + RAG logic/frontend← Next.js UI
Built with privacy and local-compute in mind. Star ⭐ if this helps you think better.
- Voice note capture (Whisper)
- Multi-agent memory (via agent.yaml)
- Docker + one-click install
- Web search tool (optional, local-first)
- Mobile PWA support
Made with ❤️ by Ayush