Justine Winata

Built a distributed video pipeline ingesting 16,000 frames per second from 2,000 CCTV cameras, with RabbitMQ retry and backpressure mechanisms for downstream failures. Now building natural-language video search over 75M detections and an incident-response agent that analyzes logs, creates Jira tickets and opens pull requests with proposed fixes.

Reduced manual QA time by 75% by building Selenium automation for a Singapore Government administration portal. Maintained 85% automated test coverage and supported a delivery process shipping more than two production deployments per day.

Built an automotive e-commerce system processing 100 product sales a month and a POS platform used to process IDR 200M+ in monthly gross sales, replacing manual workflows for an antivirus distributor.

Cut report-generation time by 96% for a production-management platform containing 16,000+ records, one of three responsive client applications delivered in six months.

Trained 100+ instructors, led academic instructor research and taught up to ten classes a semester across databases, frontend, backend, OOP, data structures, deep learning and Android.