Dr. Violet Howard demonstrates a RAG system in n8n that automatically extracts text from files, converts it to vectors, and stores them in Supabase. The pipeline watches for new or changed files, updates the vector database with metadata, and removes outdated records to keep the knowledge base current.
This new RAG AI agent is absolutely insane. It uses Postgres for memory and Supabase for the vector database. I built up this system that automatically watches for files, determines the file type, extracts the text, puts it into Supabase with a ton of metadata, and then also another system that will watch for the files are updated. It's going to delete the old records from Supabase, update the records, and also give them a new version type within Supabase. Super cool stuff. If you guys want to figure out how I built this, go to my channel and watch the video. And while you're there, check out some other AI agent tutorials. I'll see you guys over there.
