Email Classifier

In this beginner-friendly tutorial, Noah from Skillet AI shows how to create a hands-off email triage system using n8n and ChatGPT. You will learn to capture new Gmail messages, pass them to an OpenAI node for smart classification, and route each result through a three-way switch. What you’ll build Needs Reply path: star the email, generate a draft response in your own voice, and save urgency data. FYI path: summarize informational notes and log them for later review. Newsletter path: label promos, mark them read, and extract the unsubscribe link. The workflow also appends sender, subject, AI summary, draft reply, response-time estimate, and unsubscribe URL to a Google Sheet for easy tracking. Along the way you’ll refine system and user prompts, handle JSON output cleanly, convert Unix timestamps to readable dates, and test each branch with pinned data. Noah wraps up with ideas for next steps—daily digest summaries, Slack or SMS alerts, and adapting the same logic to chat or support tickets—so you can extend the project wherever automation helps most.