AI Assistant: RAG MCP & Search Engine Integration

This n8n workflow builds an advanced AI agent capable of answering user questions by accessing two specialized MCP servers: a Retrieval-Augmented Generation (RAG) MCP server for internal knowledge retrieval and a Search Engine MCP server for real-time web data. Designed for self-hosted n8n instances, it leverages community nodes to enable seamless integration, allowing the agent to provide accurate, context-aware answers combining static and dynamic information sources.

The workflow operates through a chat interface where users input queries, and the AI agent intelligently decides which MCP tool to invoke—RAG for domain-specific data or search for current events. It requires setup with an external RAG MCP server (linked template provided) and Bright Data's Web Unlocker API for search capabilities, ensuring robust handling of diverse queries without hallucinations.

Installation involves installing the n8n-nodes-mcp community node, configuring Docker for self-hosting if needed, setting up credentials, and linking endpoints. Once active, simply trigger the Chat node to interact. Benefits include time-saving automation for knowledge-intensive tasks, enhanced decision-making with hybrid data sources, and scalability for enterprise use.

Ideal for commerce and procurement: query product availability, supplier info from RAG, or market trends via search. Use cases span customer support, research, competitive analysis, and internal Q&A, delivering high-value insights efficiently.