Automated Retrieval-Augmented Generation System with Source Citations

This n8n workflow implements a comprehensive Retrieval-Augmented Generation (RAG) system that enhances data handling and AI responses by integrating Google Drive, Qdrant vector store, OpenAI embeddings, and Google Gemini. The workflow begins with creating a Qdrant collection for storing document vectors, which are generated from Google Drive documents using OpenAI embeddings. These vectors enable efficient similarity searches for relevant information. When a user submits a question, the system retrieves the most pertinent documents from the vector store via a retriever node, then uses Google Gemini to generate an answer, citing source files automatically from Google Drive. Additional nodes handle document processing, data cleaning, and response formatting, making this workflow useful for knowledge bases, customer support, or research projects requiring accurate sourced AI outputs.