RAG AI Chatbot with OpenAI, Cohere Reranker & Supabase

This advanced n8n workflow creates a powerful Retrieval-Augmented Generation (RAG) AI chatbot that answers questions based on your documents. It processes PDFs from Google Drive, extracts content, generates embeddings, and stores them in a Supabase vector database. When users query via chat, it performs semantic search, reranks results with Cohere for relevance, and generates contextual responses using OpenAI, while maintaining conversation memory.

Key benefits include highly accurate, context-aware answers without hallucinations, scalable document handling, and seamless integration of top AI services. It's ideal for customer support, internal knowledge bases, or educational tools, saving hours of manual Q&A and enabling self-service interactions.

Use cases: Deploy as a support bot for FAQs from product docs, create an internal wiki assistant for company policies, or build educational tutors from textbooks. Requires self-hosted n8n, OpenAI/Cohere/Supabase/Google Drive credentials, and a Supabase table with vector support.