As the engineer who owned data collection at Automatiq, the #1 automation platform in ticket resale with $3.5B+ in priced inventory and 500+ customers, I rebuilt the company’s core data-acquisition infrastructure from the ground up.
I inherited a small, static fleet of EC2 instances running legacy scraping code and turned it into a fully distributed system on AWS Fargate that processed over 10 million tickets daily across every major ticketing platform. Each provider got its own queue with independent autoscaling, which improved resource utilization and cut time-to-resolution for customers.
The core of the job was a cat-and-mouse game with Ticketmaster, SeatGeek, and AXS. Every few weeks a platform shipped new countermeasures: obfuscated JavaScript, browser fingerprinting, behavioral analysis, CAPTCHAs, rotating API signatures. I’d reverse-engineer the updated protections, deobfuscate the detection logic, and rebuild our integrations. That cycle ran for three years straight.
Two results worth naming. The first is the “transferless” delivery mechanism, which delivered tickets to buyers without the ticket ever leaving the seller’s account and bypassed platform transfer restrictions entirely; the fuller story is on the work page. The second is a serverless API that converted Apple Wallet ticket passes into distributable PDFs, which unlocked a category of inventory that was previously undeliverable.