Job Search Automation Pipeline
18 ranked roles out of a feed that would have taken hours to read by hand.
Job boards are built for browsing, not deciding. You run the same search every morning and the board remembers nothing about what you already rejected yesterday.
So I stopped browsing. The pipeline uses Apify’s LinkedIn scraper to pull Ontario roles, then scores them against what I actually care about. The first run surfaced 18 strong matches, ranked, out of a feed that would have taken hours to read properly.
I used an off the shelf scraper on purpose. The interesting problem is the ranking, not the plumbing, and I’ve watched enough side projects die inside the plumbing.
Next is a small personal portal with the scraping on a schedule, so the pipeline runs itself and the portal is just the morning read. The pattern generalizes to anything you’d otherwise check every day and forget.