Zac Taylor

Over the last four years, we've used, tested, and benchmarked close to 100 web scraping APIs and residential proxy providers across different websites and request volumes.

One thing has become clear to us: there is no single best proxy provider.

AI is starting to change the economics of building scrapers. A scraper that used to take a few hours of manual selector work, browser debugging, schema cleanup, retries, and maintenance can now often be scaffolded in minutes.

The part I'm interested in is the real cost, not the demo cost.

Hey hey hey

I spend a lot of time maintaining small open-source libraries and developer tools, usually after work and usually because something broke in a dependency chain.

The tools I'm most interested in are the unglamorous ones: better docs, cleaner APIs, dependency health, monitoring, retries, changelogs, and anything that saves maintainers from answering the same vague issue for the 40th time.