Diane Feddema
Best Practices Working Group Chair at Red Hat
Boulder, Colorado, United States
Summary
👤
Senior🎓
Top SchoolDiane Feddema is a Principal Software Engineer and Best Practices Working Group Chair at MLCommons with over a decade of experience optimizing performance for AI/ML systems. Based in Boulder, she specializes in tuning model training and inference across Linux and Kubernetes environments, working hands-on with accelerators like NVIDIA A100/H100 and benchmarks such as MLPerf. At Red Hat she has led performance and scale efforts for OpenShift, edge deployments, and managed services, designing experiments that uncover bottlenecks across frameworks (TensorFlow, PyTorch), storage systems, and containerized big-data stacks. Her background in high-performance computing and climate modeling informs a methodical approach to profiling and reproducible benchmarking. Known for translating complex performance data into actionable best practices, she bridges engineering, standards, and OEM partnerships to drive measurable system gains.
10 years of coding experience
27 years of employment as a software developer
Master's degree EECS Electrical Engineering and Computer Science, Master's degree EECS Electrical Engineering and Computer Science at University of Colorado BoulderBachelor of Science - BS EECS Electrical Engineering and Computer Science, Bachelor of Science - BS EECS Electrical Engineering and Computer Science at University of Iowa
Github Skills (54)
nfd8
docker8
machine-learning8
linux8
debug8
portable8
python-interface7
kmeans7
pods7
logging7
operator6
openshift6
ansible6
podman6
artifacts6
Programming languages (8)
Github contributions (5)
dfeddema/play
Aug 2016 - Aug 2018
Contributions:4 PRs, 84 pushes, 1 branch in 1 year 11 months
dfeddema/mlcube
Aug 2022 - Aug 2023
MLCube™ is a project that reduces friction for machine learning by ensuring that models are easily portable and reproducible.
Contributions:1 PR, 14 pushes, 2 branches in 1 year
fairness-mlfrictionmlmachine-learningreproducible
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.