Automated Telegram PDF Processing and Q&A Workflow

This n8n workflow automates the process of receiving PDF files via Telegram, extracting and storing their content in a Pinecone vector database, and enabling interactive question-and-answer sessions based on the stored data. Initially triggered by a new Telegram message, the workflow checks if the message contains a document, downloads the file, and converts it to PDF format if necessary. The document is then split into manageable chunks, embedded using OpenAI’s embeddings, and stored in Pinecone for efficient retrieval. When a user sends a query, the workflow retrieves relevant document chunks from Pinecone, uses a language model to generate accurate answers based on the stored data, and responds via Telegram. This setup is ideal for users who want to create an AI-powered knowledge base accessible through Telegram, such as for educational resources, project documentation, or customer support.