Document-Based Chatbot with Memory (OpenAI, Pinecone, Google Drive)

This n8n workflow creates a powerful document-based chatbot that leverages Retrieval-Augmented Generation (RAG) to answer user queries based on your private documents stored in Google Drive. It uses OpenAI for intelligent embeddings and chat completions, Pinecone as a scalable vector database for fast similarity search, and includes memory persistence via Airtable to remember user preferences even after sessions end.

Key features include a webhook-triggered chat interface for real-time conversations, automatic document loading and chunking from Drive, embedding generation, storage/retrieval from Pinecone, and an AI Agent configured for professional customer support (customizable system prompt for DGM website or similar). Benefits: Saves hours of manual data processing, provides context-aware responses reducing hallucinations, scales for production chatbots, and enhances user experience with memory for personalized interactions.

Ideal use cases: Customer support bots for websites/CMS, internal knowledge bases, FAQ handlers, or e-commerce product inquiry systems. Easy setup with your own API keys; optional Airtable removal for session-only memory. High time-saving potential for non-technical users starting with AI chatbots.