The internet holds a huge amount of valuable data—and to us as engineers it’s honestly beautiful. In this video I show you how to harness that data at industrial scale using Node.js, Puppeteer, and Bright Data’s Scraping Browser. The goal is simple (and it’s my biggest pet peeve): job ads that hide salary info. So we build a small app that scrapes indeed.com, extracts relevant job details, and only returns listings that match your criteria and actually include a salary. I walk through spinning up a Next.js project, refactoring the folder structure the way I like it, and wiring an API route that runs the scraper. Then we connect Puppeteer (puppeteer-core) to Bright Data via a browser WebSocket endpoint, so we can avoid the usual scraping pain: IP bans, bot detection, and captchas killing your workflow. You’ll see the scraper fill a CSV in real time, scale from small runs to hundreds/thousands of results, and then apply filtering (like requiring a £ sign) so “full-time” doesn’t pretend to be a salary. To finish, I show my favorite little trick: using ChatGPT to write Puppeteer extraction code. You copy an element’s HTML from DevTools, paste it into ChatGPT, and it can generate the selector logic for you—saving you from staring at nausea-inducing HTML when you just want the data.