Self‑hosted AI‑powered markdown knowledge base with hybrid search
It lets teams turn collections of markdown documentation into an interactive, AI‑driven knowledge base. By indexing docs with both BM25 keyword vectors and semantic embeddings, it offers fast hybrid search and a chat interface that answers queries using local or cloud LLMs. The project provides a FastAPI backend, a React frontend, Docker deployment, CLI tools and a plugin system, making setup easy for developers needing searchable docs without SaaS lock‑in. Compared to generic vector search tools, it adds versioned history, knowledge‑graph links and a planner for richer documentation context.
View on GitHub →Z7Lab/markdownkb