Built a real-time warehouse operations dashboard with live forklift tracking and WebSocket event streaming. Presented weekly demos to stakeholders, translating technical architecture into clear operational value for both engineering and operations teams.

Developed and documented backend APIs for a hospital management system, then traced and resolved a performance bottleneck that cut page load time from 2.5s to 1.5s. Orchestrated a full server migration, reconfiguring Nginx, Redis, RabbitMQ, and database services on new infrastructure.

Built a food-recognition AI system achieving 96.5% accuracy across 201 categories, exceeding the CDC's 90% target. Deployed a live demo on cloud infrastructure for stakeholder evaluation, then re-architected the model for 2× faster inference and presented the accuracy-speed trade-offs to guide the final production decision.