We run a B2B SaaS platform serving specialty running retailers. Part of what we do is pull inventory data from brand-side wholesale portals so our retailers can see live vendor availability alongside their own POS data. We have valid B2B accounts on these portals. We have around 15 of these scrapers running today and want to add more, plus we need someone reliable to fix existing ones when a vendor changes their site.
What you'll build
• New scrapers for brand B2B portals to extract inventory data (SKU, size/color/width, quantity, backorder dates). Output goes into a MySQL 8 schema we'll provide.
• Product image scrapers for both B2B portals and public-facing brand sites.
• Fixes and updates to existing scripts when a portal changes (auth flow updates, endpoint changes, HTML/JSON structure shifts).
Stack
PHP 8.4 is our primary language. Existing scrapers are based on PHP, Selenium and Python. Use whichever HTTP library you find cleanest. Selenium, Playwright, or Puppeteer are welcome when a portal has complex JS-driven rendering or session flows that don't work with straight HTTP. Node or Python is fine for those. Scripts run on a Ubuntu server on GCP.
Requirements
• Comfortable inspecting network traffic to figure out how an undocumented portal actually works behind the login page. Most brand portals have a JSON API driving their UI even when it isn't publicly documented.
• Willing to do both greenfield builds and maintenance on existing scripts.
Engagement
Hourly and ongoing. Volume varies week to week. Several existing scripts need to be reworked and occasionally a new script needs to be created for a new brand. When an existing breaks, it's usually because a vendor changed their portal without notice, and we need to be able to reach you promptly. We're prioritizing quality and a successful outcome over cost. If you're the right fit, we're not looking to bid you down.
To apply
Share one or two examples of authenticated portal automation you've built, ideally something involving JWT or cookie/session auth.