RAG Chatbot: Google Drive Docs to GPT-5 & Supabase Vector DB

This advanced n8n workflow creates a complete RAG (Retrieval-Augmented Generation) pipeline for transforming documents from Google Drive into a searchable AI knowledge base. New files in a designated folder are automatically detected, processed with Mistral AI for OCR text extraction (including tables), intelligently chunked (one page per chunk for optimal context), embedded using OpenAI models, and stored in Supabase vector database for semantic search.

When querying, the workflow retrieves relevant chunks, augments prompts with GPT-5 (OpenAI's latest model) for precise, context-aware answers, and maintains conversation history in Postgres for coherent, multi-turn chats. Benefits include instant access to document insights without manual searching, reduced research time from hours to seconds, and scalable knowledge management.

Ideal use cases: Enterprise knowledge bases for legal/contract review, customer support FAQs from manuals, research teams analyzing reports/PDFs, content creators querying media libraries, or any scenario needing AI-powered document Q&A. Unlock hidden value in your files with enterprise-grade automation.