
This AI research agent is an automated n8n workflow that fetches the latest content from RSS feeds (YouTube/Blogs), filters for fresh news (last 24h), and uses Google Gemini Pro 1.5 and SerpAPI to generate structured podcast topic ideas. The results are automatically saved to a Notion database and sent to Telegram as a notification.
Based on my custom n8n template, the workflow is broken down into three logical blocks:
Specialist Tip: I’ve included a filter to exclude "YouTube Shorts" to ensure we only focus on deep-dive content for podcast research.
The Difference Maker: I’ve connected SerpAPI to the AI Agent. This allows the model to perform real-time Google searches on trending topics, ensuring the content ideas aren't just creative, but factually current.
Structured Output: Instead of a wall of text, the agent is forced to output JSON data with specific fields: Podcast Title, Topical/Contrarian angle, Hook, and Thumbnail Idea.
The Final Touch: To keep the team agile, the workflow sends a summary message via Telegram Bot, alerting you that your "Daily Research" is ready to review in Notion.
As an automation specialist, I use n8n because it provides more flexibility than Zapier for complex loops and structured data handling. For example, handling multiple RSS feeds at once is significantly more efficient in this environment.
I’ve made this entire workflow available as a free template. You can import it into your n8n instance, connect your Google Sheet and Notion account, and start automating your research today.