Dynamic RAG AI Agent with Supabase, PostgreSQL, Google Drive in n8n

This quick tutorial shows how to build a dynamic Retrieval-Augmented Generation (RAG) AI agent in n8n that keeps your website product details and organization information always up to date. It uses PostgreSQL and Supabase for vector storage, and Google Drive for managing documents dynamically—add, update, or delete files to instantly refresh your knowledge base.

Project Objective

The goal is to create a smart RAG agent that pulls accurate, real-time context from your dynamic knowledge base. It updates product and company details automatically whenever files change in Google Drive—no manual updates needed.

Workflow Overview

• Connect Supabase and PostgreSQL to store and manage vector embeddings
• Use n8n to monitor Google Drive for new, updated, or deleted documents
• Automatically generate, update, or remove embeddings to sync with changes
• Perform vector search to retrieve the latest context for product and organization queries
• Use OpenAI to generate accurate answers using up-to-date context
• Maintain conversation memory for better user interactions

Key Features

• Dynamic vector storage with PostgreSQL + Supabase
• Google Drive integration for real-time file intake and removal
• Automatic update and delete logic for embeddings
• OpenAI for high-quality generation with current context
• Session memory for continuous conversations
• Fully no-code orchestration using n8n

Use Cases

• Dynamic website content agents for product updates
• Knowledge base bots for internal teams
• Customer support with real-time product and company info

Results

With Supabase, PostgreSQL, Google Drive, and OpenAI running in n8n, this agent can update your website knowledge base automatically—add new product details, modify existing info, or remove outdated data without any backend code changes.