Conversational rag

★ 7.1 · n8n · medium · 7 nodes

Conversational RAG is an n8n automation template for building a conversational chatbot that answers questions over your own documents using Retrieval-Augmented Generation. Reach for it when you need an internal Q&A assistant over a corporate knowledge base, when users should ask questions against uploaded PDFs or text files with full conversation memory, or when you want a working RAG setup without building the pipeline from scratch. The template ships as three linked workflows: document upload and indexing into a Qdrant vector store, a conversational RAG agent with session memory, and an error handler. Trigger: chat. Integrations: OpenAI (LLM), Qdrant (vector store). Deployed by importing JSON workflow files into a self-hosted or cloud n8n instance — no custom code required. The direction is retrieval-grounded Q&A over provided documents, not open-ended generation or summarisation without a source. Suitable for teams that already have n8n, an OpenAI API key, and a Qdrant instance; not a fit if fully offline or API-free operation is required.