Full RAG Tutorial | AI Chat with YouTube | LangChain, Node, Supabase, Postgres, pgvector, React

In this video, we'll build a RAG app for chatting with YouTube videos. You will be able to paste a YouTube URL and ask literally anything about it (and have a continuous chat). We have 3 parts: 1. Frontend (React + Vite) 2. Backend (Supabase + PostgreSQL + pgvector) 3. RAG service (Node + Langchain) Works pretty fast; even without further optimization we get the entire transcript, process it for AI, and store it in a vector database in ~ 5 seconds. Then, every time we ask a prompt, it takes 2-3 seconds to start streaming. All the knowledge comes from building Second Brain. Chat with YouTube videos and way more file sources, build your virtual memory, AI chat with it, and way more. Check it out https://www.thesecondbrain.io/?utm_source=youtube&utm_medium=social&utm_campaign=full_rag_tutorial