I work on display ads across the Prime Video homepage. The goal is fast, reliable delivery at scale; in practice that means services in Kotlin/Java, latency-critical paths and custom TPS generators written in C++, and Python for AWS CDK/IaC. I focus on turning ambiguous problems into measurable improvements—safe traffic replay and shaping, p90/p99 tuning, and calm multi-region rollouts.

Built a C++ CLI that summarizes large files to speed up how analysts and consultants dig through data.

Built a full-stack application and a tree-node data visualization tool to display key safety performance indicators for autonomous vehicles.

Conducted research experiments under PhD student Jack Ma and was supervised by Karl Herrup (Stanford PhD), Head of the Alzheimer’s Disease Research Center at the University of Pittsburgh School of Medicine. I helped with experimental design, data collection, and computational analysis to surface findings for the broader team.