Austin, Texas Metropolitan Area
Infrastructure Line of Business
• Developed and led project team for an application to autonomously run and queue benchmarks enabling faster performance insights
• Develop, improve, and analyze microbenchmarks in C to improve future IP by finding microarchitectural bottlenecks and issues
• Investigate and analyze workload performance, power usage, and software bottlenecks using hardware performance counters, eBPF, and statistics via Pandas across a wide variety of workloads, including C, Python, and Java based workloads
• Created a thread-safe and extendable eBPF script to record function statistics and usage so that IP, Compiler, and Kernel teams can optimize their products using real-world data
• Constructed a scheduling service with Python, PHP, and MAAS so that employees can reserve and use server machines while increasing total machine availability
• Implemented timeseries collection and graphing infrastructure with Python, Pandas, and Matplotlib to record, display, and compare performance output over time
• Build toolchains and kernels to test new kernel and compiler features and relay performance improvements/degradations