Joseph McRell

I build and maintain scrapers for public records and procurement data - the fragmented, unglamorous sources that don't have APIs, and that most scrapers quietly stop working against.

Every Actor I publish is checked against its live target every night by an automated monitor. When a site changes its layout, the monitor re-derives the selectors, validates them against two separate pages and a schema contract, and ships the fix only if it passes. The goal is boring reliability: if a source moves, the repair usually lands before you notice anything broke.

Currently focused on US state business registries and procurement portals.