Private Document Q&A with Llama3, Postgres, Qdrant & Google Drive

LocalRAG.AI is a fully private, on-premises AI assistant designed for handling confidential documents. It automatically monitors specified Google Drive folders for new or updated files, downloads them, extracts text content, and generates embeddings using a local Ollama instance running Llama3. These embeddings are then stored in Qdrant, a local vector database, enabling efficient semantic search.

During interactions via a webhook or chat trigger, the system performs vector similarity search in Qdrant to retrieve relevant document chunks. It combines these with conversation history fetched from Postgres, feeding them into a LangChain AI agent powered by the local Llama3 model to generate accurate, context-aware responses. This ensures complete data privacy as nothing leaves your infrastructure.

Benefits include maximum data security for sensitive info, no reliance on cloud AI services like OpenAI, cost savings from local models, and scalability for internal knowledge bases. Ideal use cases: research teams with proprietary data, enterprise internal docs Q&A, secure AI chatbots for compliance-heavy industries like finance or healthcare, and offline-capable knowledge retrieval systems.

Requires self-hosted n8n, Ollama, Qdrant, and Postgres. Setup involves authenticating Google Drive, configuring credentials, and importing the workflow. Perfect for teams prioritizing privacy over cloud convenience.