GemNews

All stories · AI Workflows

AI Ads Intelligence System

Turn an AI agent into an ad-research tool: a total beginner's workflow

By Iris (Nghi Chau) ·

Share: WhatsApp · Telegram · X · Facebook · LinkedIn · Email

Researching competitor ads often means opening ads one by one, copying details into a spreadsheet, manually tagging each creative, and starting over again the next time.

This session shows you how to turn that repetitive process into an AI-powered workflow using Claude Code, Playwright, and live ad-library data.

You’ll learn how to:

  • Collect live ads, creatives, copy, start dates, and destination links
  • Clean and organize the data into fields you can compare
  • Use AI to classify hooks, angles, formats, and landing-page patterns
  • Build a local dashboard where you can filter, group, and explore the results
  • Deploy the tool with a database and scheduled crawler so it can track what changes over time

By the end, you’ll understand how to build two versions of the workflow: a simple research tool that runs locally, and a deployed version with a public URL, database, and daily updates.

More importantly, Ads Intelligence is only one example. The same workflow can be reused for many ecommerce research tasks: collect the data, clean it, classify it, compare it, and turn it into something your team can use again.

Discuss on GemCommunity