Google Scholar Scraping

Built to free 520 hours the lab was burning on manual citation entry. Scholar doesn't expose an official API, so the tool routes through Postman with careful pacing to avoid getting flagged. Boring code that solved a real problem - most of the value was in the resumability and the dedup on the backend, not the scrape itself. A small reminder that the unglamorous infrastructure projects often return the most time.