Chat with Your Documents on Telegram via Gemini

# 🤖 Telegram AI Assistant for Your Documents (n8n + Supabase + Gemini) This project transforms a standard Telegram bot into your dedicated AI assistant – designed to understand and answer questions based on your own documents. It seamlessly integrates the power of Google Gemini for advanced language capabilities and Supabase's vector database for efficient, intelligent document retrieval. Built entirely within the no-code platform n8n, it allows you to deploy a sophisticated document chatbot without writing a single line of code. Simply upload any PDF document to the bot, and instantly gain the ability to chat with it, querying its contents as if it were a knowledgeable expert on your uploaded files. --- ## 📹 Watch the Bot in Action - Users pose questions directly to the Telegram bot. - The bot generates relevant, informative answers using the cutting-edge capabilities of the Google Gemini LLM. - Leveraging a powerful vector search mechanism, it can pull specific, contextual information from previously uploaded documents to provide highly relevant and informed responses. - (Optional) Augment answers with real-time data, like current weather information. ### 2. Effortless Document Integration (User Upload PDF → Processing → Searchable Knowledge) - Users upload a PDF document directly to the bot. - The workflow automatically parses the document content, converts it into numerical representations called embeddings using Gemini's embedding models. - These embeddings, alongside the document's text content, are then securely stored in a dedicated Supabase vector table, creating a searchable knowledge base. - Immediately after successful processing, the document becomes part of the bot's memory, enabling users to ask questions about its contents via the standard chat interface. --- ## 🧠 Core Intelligence Features - ✅ Pure No-Code: Developed and managed entirely within the intuitive [n8n: Employs Supabase as a robust vector database for storing and efficiently searching document embeddings, providing the bot with long-term memory about your content. - ⚡️ Rapid & Private Retrieval: The vector search allows for swift identification and retrieval of the most relevant document snippets based on the user's query. This approach enhances response speed and significantly improves data privacy, as the original document content remains securely stored in your Supabase instanc