🤖 AI Powered RAG Chatbot for Your Docs + Google Drive + Gemini + Qdrant

AI-Powered RAG Chatbot for Google Drive & Qdrant This n8n workflow creates a comprehensive, two-part system for building and interacting with a Retrieval-Augmented Generation (RAG) chatbot. The first part handles the ingestion, processing, and storage of documents from Google Drive into a Qdrant vector database. The second part provides a chat interface that uses AI to answer questions based on the content of those stored documents. Technologies Used - n8n: The core automation platform coordinating the entire process. - Google Drive & Google Docs: Used as the source for documents and for logging chat history. - Qdrant: A vector database used to store document embeddings for efficient similarity searches. - Google Gemini: The Large Language Model (LLM) used for AI-powered metadata extraction and for generating conversational responses in the chatbot. - OpenAI: Used to generate the text-embedding-3-large vector embeddings that represent the documents' semantic meaning. - Telegram: Integrated for sending notifications and for implementing a "human-in-the-loop" approval step for critical operations. What It Automates This workflow automates the end-to-end process of creating a knowledge base from a set of documents and making it conversationally accessible. - Document Ingestion Pipeline: It automates finding new documents in a Google Drive folder, downloading them, extracting their text, breaking them into chunks, and generating vector embeddings. - Metadata Enrichment: It uses AI to automatically read each document and extract key metadata (like themes, keywords, and conclusions), which dramatically improves search relevance. - Vector Database Management: It handles the entire "upsert" process into Qdrant. It also includes a secure, human-verified workflow for deleting specific documents from the vector store, preventing accidental data loss. - Conversational AI Chatbot: It provides a ready-to-use chat interface that can understand user questions, retrieve relevant information from the knowledge base, and generate human-like answers.