Extract structured data from any website in 2 minutes

Web scraping usually means wrestling with Playwright, rotating proxies, and getting blocked by Cloudflare. Browse AI makes it feel like using a highlighter.

Go to browse.ai and create an account. Free tier: 1 robot, 50 runs/month, enough for this lesson. No card required.

From the dashboard click Build Robot. A panel asks for a starting URL. Paste in any Amazon product page — pick something you actually care about, like a kitchen gadget or a book. Hit Record.

Browse AI opens a real browser inside your browser (it's a Chrome instance running in their cloud). You'll see the product page load exactly as a human sees it — JavaScript rendered, images loaded, everything.

Click the Capture Text button in the sidebar, then click the product title on the page. A dialog opens: name this field `title`. Done. Click Capture Text again and click the price — name it `price`. Again for the star rating — name it `rating`. Do the same for review count.

You've just taught a robot to read that page layout. Click Finish Recording at the top right. Name your robot `Amazon Product Watcher`.

Back on the dashboard, click your new robot and hit Run. In about 15 seconds you get a row of structured JSON: `{"title": "...", "price": "$29.99", "rating": "4.6", "review_count": "1,203"}`.

No code. No proxies. No Cloudflare drama. The robot used a real browser fingerprint, handled the page's JavaScript, and extracted clean data.

You can run this robot 50 times on the free tier. What you can't do: schedule it to run automatically every morning, scrape lists (not just single products), watch for changes, or send results to Slack/Sheets. That's what Starter tier unlocks — and that's what the rest of this course builds into a real business tool.