I built this n8n workflow that implements a Retrieval-Augmented Generation (RAG) AI Agent designed to answer queries grounded in private company documents without hallucinating.
Trigger: Initiated when a user chat message is received.
AI Agent Core: Coordinates the response using an LLM backbone.
Google Gemini Chat Model & Simple Memory: Powers conversational logic and tracks dialogue context across the session.
Qdrant Vector Store & Embeddings (Gemini): Retrieves relevant context from indexed internal documents.
0
132
