YouTube Comments Sentiment & Keywords Analysis with Gemini AI & Telegram

This workflow automates the end-to-end analysis of YouTube video comments. It starts by pulling a list of videos from a Google Sheets template, retrieves up to 100 top-level comments per video via the YouTube Data API, and parses them into individual items. Each comment is then processed by Gemini AI (via OpenRouter) to extract sentiment (positive/neutral/negative with score -1 to 1), cleaned keywords (stopwords/emojis/hashtags removed), and language code.

Results are normalized for consistency and written back to Google Sheets alongside original comment data, ensuring no duplicates via comment_id tracking. An aggregate summary is generated, highlighting key insights like overall sentiment and top keywords, then sent as a concise report to a specified Telegram chat.

Ideal for content creators, marketers, and analysts monitoring audience feedback. Benefits include time-saving automation, scalable analysis across multiple videos, actionable insights without manual effort, and seamless integrations. Runs on schedule for ongoing monitoring, with easy setup using standard API credentials.