I am a Senior CPU/GPU Design & Verification Engineer working on large-scale, high-performance computing platforms that power modern AI and scientific workloads. My work focuses on full-chip coherency, heterogeneous compute systems, and understanding how Python-based machine learning frameworks behave across different GPU architectures, memory hierarchies, and SDK environments. I spend much of my time investigating nondeterministic behavior, performance drift, and reproducibility issues in real-world AI pipelines.
Alongside my engineering career, I am an active researcher with published work on AI–hardware interaction, cache-coherency behavior, logic locking, and cross-SDK GPU benchmarking. I enjoy bridging hardware concepts with practical Python workflows and helping developers understand how underlying system behavior influences model training and inference.
Previously, I served as a Teaching Assistant at the University of Southern California while completing my master’s degree. I also mentor students, judge hackathons, and frequently speak at technical conferences about AI systems, performance engineering, and debugging complex workloads.
I am passionate about helping Python developers build stable, reliable, and reproducible AI pipelines by demystifying the layers of software and hardware that influence real-world model behavior.