Haiku RAG
Agentic RAG system powered by LanceDB, Pydantic AI, and Docling
Retrieval-augmented generation system providing hybrid search combining vector and full-text search with Reciprocal Rank Fusion. Features multi-agent research workflows for planning, searching, evaluating, and synthesizing information with citations. Supports 40+ file formats including PDF, DOCX, HTML, and Markdown. Works locally with embedded LanceDB or supports cloud storage on S3, GCS, and Azure.
Source
Repository: https://github.com/ggozad/haiku.rag
Maintain Haiku RAG?
Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.
[](https://getagentictools.com/mcp/ggozad-haiku-rag?ref=badge) Related mcp
Redis Provides a natural language interface to Redis databases, enabling operations on various data structures with tools for vector sim
MotherDuck & DuckDB Integrates MotherDuck and local DuckDB databases for flexible querying and analysis of structured data in MCP-compatible environme
Second Brain Self-hosted semantic memory MCP server on Cloudflare Workers with vector embeddings for intelligent recall and 4 tools for persona
Docfork Retrieves up-to-date documentation and code examples for any software library through the Docfork API, automatically selecting rel
Chroma Integrates with Chroma vector database to enable collection management, document operations, and vector search capabilities for kn
WeRead Integrates with WeRead e-book platform to provide access to book content, user highlights, and annotations with efficient cache ma