Phong Cao

Optimized a C++/CUDA inference pipeline, increasing GPU throughput by 28% and reducing p95 latency by 21%. Containerized GPU-backed services and standardized Kubernetes deployment checks, cutting release time from 45 to 18 minutes.

Developed Python services and Angular dashboards for telemetry workflows processing 2.5 million events per day, reducing dashboard load time from 3.8 to 1.9 seconds. Evaluated anomaly-detection workflows, achieving 91% precision and reducing false alerts by 24%.

Built Python and Java APIs sustaining 18,000 requests per second, decreasing p99 response time by 35% through caching and parallel processing. Improved ML-assisted ranking by 12% and reduced manual review volume by 30%.

Created Python and C++ network-automation tools validating configurations across 10,000+ devices, cutting validation cycle from 90 to 24 minutes. Optimized SQL-backed telemetry services handling 400 million records monthly, reducing query latency by 55%.

Conducted Python-based research experiments on large datasets, reducing preprocessing time by 42% via vectorized transformations. Evaluated machine-learning algorithms and developed a Python and Angular departmental application.

Built a strong foundation in algorithms, data structures, databases, and software engineering. Applied classroom concepts through research, web development, and hands-on programming projects.