🧠 Supabase RAG Agent using n8n and supabase

This project implements a RAG (Retrieval-Augmented Generation) pipeline using n8n, integrated with OpenAI, Supabase, and PostgreSQL. It enables contextual AI chat with persistent memory and document search capabilities.

📦 Features

• 🧠 RAG Agent powered by OpenAI's GPT model
• 💾 Postgres-backed chat memory for context retention
• 📚 Supabase vector store for embedding search
• 📂 Google Drive integration to load documents
• 🧩 LangChain nodes for document parsing and splitting
• 🔄 Automated embedding + indexing pipeline

🛠️ Components

• OpenAI Chat Model: Handles user interaction via GPT-4o-mini
• Postgres Memory: Stores past messages for contextual understanding
• Supabase Vector Store: Enables semantic search with embedded documents
• Document Loader & Splitter: Parses binary data, splits text for embedding
• Google Drive Node: Automatically pulls documents from Drive

🚀 Workflow Overview

1. Trigger: Chat message or manual trigger
2. RAG Agent uses OpenAI + Postgres Memory + Supabase Vector Search
3. Binary files from Drive are loaded, split, embedded
4. Embeddings stored in Supabase for retrieval
5. Answers are returned with contextual + factual relevance

📂 How to Use

1. Download the workflow and auth it .
2. Connect credentials for:
   • Google Drive
   • OpenAI
   • Supabase
   • Postgres
3. Upload documents to the specified Drive folder.
4. Trigger the workflow using chat or manual test.
5. Enjoy your intelligent assistant with memory and vector search!