In this video I show you how I build a super powerful RAG AI agent with n8n + Supabase in basically minutes. Before, setting up RAG used to take a lot of time and effort, but with the newer n8n updates and the way Supabase works as a vector store, the workflow is insanely simple: upload your docs once, then let the agent retrieve answers from your own database instead of hallucinating. I walk you through the exact setup: creating a free Supabase account, connecting credentials inside n8n, and using the Supabase Vector Store node to insert documents. Then I connect the same vector store as a tool inside my n8n AI Agent so it can retrieve documents at chat time. I show the full flow with a real example: I upload a Word doc from Google Drive, generate embeddings (I use OpenAI embeddings), and then ask the agent a question like “what services do you offer?”—and it pulls the answer directly from the document. The big takeaway is you can package this for businesses with lots of PDFs/Docs and charge real money (I’m talking $1k–$5k builds plus a monthly retainer) because it saves them hours of digging through old files.