RAG Agent

User questions are first matched with relevant document sections in a vector database, and the AI then generates an answer using that context.

Key Features

• Document-based AI question answering
• Supabase Vector Store for semantic search
• HuggingFace embeddings for document vectorization
• Automated document ingestion pipeline

Workflow Architecture

Tech Stack

• N8N
• Supabase (Vector Store)
• HuggingFace Embeddings-Embedding model
• Google Drive
• Simple Memory (n8n node)

Workflow Loom Video Demo ——->
Screenshot of Workflow ———> ScreenShot