Automate a 'Chat with your PDF' Bot on Telegram with Google Gemini & Pinecone

This n8n template from Intuz provides a complete solution to automate a powerful, AI-driven 'Chat with your PDF' bot on Telegram.

It uses Retrieval-Augmented Generation (RAG) to allow users to upload documents, which are then indexed into a vector database, enabling the bot to answer questions based only on the provided content.

Who's this workflow for?

• Researchers & Students
• Legal & Compliance Teams
• Business Analysts & Financial Advisors
• Anyone needing to quickly find information within large documents

How it works

This workflow has two primary functions: indexing a new document and answering questions about it.

1. Uploading & Indexing a Document:

• A user sends a PDF file to the Telegram bot.
• n8n downloads the document, extracts the text, and splits it into small, manageable chunks.
• Using Google Gemini, each text chunk is converted into a numerical representation (an "embedding").
• These embeddings are stored in a Pinecone vector database, making the document's content searchable.
• The bot sends a confirmation message to the user that the document has been successfully saved.

2. Asking a Question (RAG):

• A user sends a regular text message (a question) to the bot.
• n8n converts the user's question into an embedding using Google Gemini.
• It then searches the Pinecone database to find the most relevant text chunks from the uploaded PDF that match the question.
• These relevant chunks (the "context") are sent to the Gemini chat model along with the original question.
• Gemini generates a new, accurate answer based only on the provided context and sends it back to the user in Telegram.

Key Requirements to Use This Template

1. n8n Instance & Required Nodes:

• An active n8n account (Cloud or self-hosted).
• This workflow uses the official n8n LangChain integration (@n8n/n8n-nodes-langchain). If you are using a self-hosted version of n8n, please ensure this package is installed.

2. Telegram Account:

• A Telegram bot created via the BotFather, along with its API token.

3. Google Gemini AI Account:

• A Google Cloud account with the Vertex AI API enabled and an associated API Key.

4. Pinecone Account:

• A Pinecone account with an API key.
• You must have a vector index created in Pinecone. For use with Google Gemini's embedding-001 model, the index must be configured with 768 dimensions.

Setup Instructions

1. Telegram Configuration:

• In the "Telegram Message Trigger" node, create a new credential and add your Telegram bot's API token.
• Do the same for the "Telegram Response" and "Telegram Response about Database" nodes.

2. Pinecone Configuration:

• In both "Pinecone Vector Store" nodes, create a new credential and add your Pinecone API key.
• In the "Index" field of both nodes, enter the name of your pre-configured Pinecone index (e.g., telegram).

3. Google Gemini Configuration:

• In all three Google Gemini nodes (Embeddings Google Gemini, Embeddings Google Gemini1, and Google Gemini Chat Model), create a new credential and add your Google Gemini (Palm) API key.

4. Activate and Use:

• Save the workflow and toggle the "Active" switch to ON.
• To use: First, send a PDF document to your bot. Wait for the confirmation message. Then, you can start asking questions about the content of that PDF.

Connect with us

• Website: https://www.intuz.com/services
• Email: [email protected]
• LinkedIn: https://www.linkedin.com/company/intuz
• Get Started: https://n8n.partnerlinks.io/intuz

For Custom Workflow Automation

Click here- Get Started