watsonx-rag-mcp-server

The tutorial guides you through creating a Retrieval-Augmented Generation (RAG) server using IBM Watsonx.ai, ChromaDB for vector indexing, and the Model Context Protocol (MCP) Python SDK. The server processes PDF documents and allows clients to query them, leveraging large language models grounded in specific data. The server is designed to be standalone, with a client for making queries. The tutorial also covers integrating the server with Claude Desktop, providing a seamless interface for interaction.

Features

• Integration with IBM Watsonx.ai for LLM capabilities.
• Use of ChromaDB for efficient vector indexing and retrieval.
• Implementation of the Model Context Protocol (MCP) for standardized communication.
• Ability to process and query PDF documents for specific information.
• Integration with Claude Desktop for direct interaction.

MCP Tools

• {'chat_with_manual': 'A tool that answers questions about drone manuals using RAG with Watsonx.ai.'}

Usage with Different Platforms

mcp

{'command': 'python', 'args': ['server.py'], 'workingDirectory': '/path/to/your_project'}

docker

docker run --rm -p 5001:5001 -v "$(pwd)/.env:/app/.env" -v "$(pwd)/documents:/app/documents" -v "$(pwd)/chroma_db_data:/app/chroma_db_data" --name rag-chat-container ruslanmv/watsonx-rag-chatbot:latest