Building OpenBrain: A Personal RAG Knowledge System

Full write-up coming soon.

OpenBrain is a personal knowledge system built on RAG (Retrieval-Augmented Generation). The core idea: instead of re-explaining context at the start of every AI session, ingest it once and query it on demand.

Topics covered in this session:

• Pinecone vector store setup and namespace strategy
• Vercel deployment and MCP server integration
• Claude Code session hooks for automatic context injection
• Iterating on the ingest format to improve retrieval quality