n8n Docs Expert Chatbot with OpenAI RAG Pipeline

This workflow creates a Retrieval-Augmented Generation (RAG) system turning an AI into an n8n documentation expert. Part 1 handles one-time indexing: it scrapes all n8n docs pages, splits them into chunks, generates embeddings with an AI model, and stores them in n8n's Simple Vector Store. This builds a searchable knowledge base like detailed index cards for every doc section.

Part 2 is the interactive chatbot. On query, it retrieves the most relevant chunks from the vector store, feeds them to a Gemini model with instructions to answer solely from that context, ensuring factual, precise responses without hallucinations. The in-memory store is temporary, requiring re-indexing after n8n restarts.

Benefits include hyper-accurate answers for complex n8n queries, easy adaptation to other docs, and no external vector DB needed. Ideal for n8n power users, support teams, or learning automation—setup in 2 mins, indexing in 15-20 mins. Use cases: internal docs Q&A, product expertise bots, or custom knowledge assistants.