Chatbot Gemini Video Analysis - Automated video analysis

Process description

This automation is a chatbot that summarizes YouTube videos. It receives requests via Telegram, analyzes the videos provided using Google Gemini, and sends a detailed resume back to the user. The resume, if it is in another language, is translated into Russian. The response size limit is 600 characters to avoid sending a message that is too long.

API keys and services

• Telegram API — to interact with a chatbot.
• Google Gemini (PalM) API — for video analysis.

System architecture by blocks

SECTION 1: RECEIVING A REQUEST

1. Telegram Trigger - Receiving a request
   
   • Purpose: Automatically receives messages from a user on Telegram.
   • Settings: Updates are tracked by the 'message' type.
   • What we get: The text of the message from the user.

SECTION 2: ANALYSIS AND PROCESSING OF INFORMATION

1. AI Agent
   
   • Purpose: The main block for processing the request. Acts as an AI agent that summarizes videos.
   • Settings:
     
     • Prompt: You are an AI agent for summarizing YouTube videos. Describe the entire video in detail. If the description is in another language, then translate it into Russian. 600 characters limit, don't write more than one message.
     • Input data: The text of the message from the user.
   • Connected components:
     
     • Google Gemini Chat Model — language model for chat processing.
     • Simple Memory — for storing chat history.
     • Analyze video in Google Gemini — video analysis tool.
2. Analyze video in Google Gemini
   
   • Purpose: A video analysis tool.
   • Settings:
     
     • Resource: 'video'.
     • Operation: 'analyze'.
     • Model: 'models/gemini-2.5-flash'.
     • Input data: The URL of the video you want to analyze.

SECTION 3: SENDING A RESPONSE

1. Send a text message
   
   • Purpose: Sends the video analysis result back to the Telegram user.
   • Settings:
     
     • Chat ID: The chat ID from which the original request came.
     • Text: The result obtained from AI Agent.

Node connection diagram

• Main stream:
  
  1. Telegram Trigger → AI Agent.
  2. AI Agent → Send a text message.
• AI Connections:
  
  1. Google Gemini Chat Model → AI Agent (language model).
  2. Simple Memory → AI Agent (chat memory).
  3. Analyze video in Google Gemini → AI Agent (video analysis tool).

System capabilities

• Supported features: Analyzing and summarizing YouTube videos.
• Language support: Translation of the video description into Russian, if the original description is in another language.
• Response format: The answer is limited to 600 characters.
• Automated processes:
  
  • Automatic receipt of requests from Telegram.
  • Automated video analysis.
  • Automatically send a resume to the user.

System application

• For users: Quickly get summaries of long videos without having to watch them.
• For content creators: Quick analysis of competitors or your own videos.

The result of the system

• What happens is: Instant YouTube video resume.
• Save time: A significant saving of time that could be spent watching the entire video.
• Convenience: A convenient and easy way to get information directly from Telegram.