Qdrant Hybrid
bymarlian
0Offers a local-first TypeScript MCP server for Qdrant, facilitating client isolation, LM Studio integration, and scalable document workflows for retrieval-augmented generation.
About
This tool delivers the ultimate evolution of Retrieval-Augmented Generation (RAG) systems by combining architectural best practices and performance optimizations from `lance-mcp`, `sqlite-vss-mcp`, and `delorenj/mcp-qdrant-memory`. The result is a production-ready, multi-tenant RAG system that provides complete client isolation, advanced document seeding, and seamless integration with LM Studio for fully local LLM operations, ensuring maximum privacy and enterprise-grade document processing capabilities.
Key Features
- Enterprise-grade semantic and granular search capabilities across clients and documents
- Multi-client architecture with complete data isolation and separate collections
- Seamless LM Studio integration for local BGE-M3 embeddings and Qwen3 summaries
- Advanced document processing with SHA256 deduplication and multi-format support (PDF, MD, TXT, DOCX)
- Local-first design ensuring maximum privacy and zero cloud dependency
- 0 GitHub stars
Use Cases
- Building enterprise-grade semantic search capabilities for diverse document formats
- Managing RAG systems for multiple clients or projects with complete data isolation
- Deploying a privacy-focused RAG system for sensitive, local documents
Sponsored