Build a Secure Document Q&A System with Llama3, Postgres, Qdrant, and Google Drive

This workflow is designed for self-hosted n8n instances, enabling you to create a secure and private AI assistant that uses your own documents to answer questions. By monitoring specified Google Drive folders, the system downloads and processes new or updated files, extracts text, and generates embeddings using a local Llama3 model. These embeddings are stored in a Qdrant vector database. During interactions, the system retrieves relevant information using vector search, combines it with chat history stored in Postgres, and responds through a LangChain AI agent. This setup is ideal for research teams handling confidential data and internal documentation Q&A, without relying on cloud services.