Garrett Goon
Research Scientist at IBM
Greater Philadelphia United States
Summary
🤩
Rockstar🎓
Top SchoolGarrett Goon is a research scientist and senior AI/ML engineer with seven years of experience bridging cutting-edge theoretical physics and production-scale machine learning. Trained as a theoretical physicist with a PhD from Penn and ~20 papers cited ~1,000 times, he pivoted into ML engineering at Determined AI/HPE and now IBM, specializing in PyTorch and distributed training at scale. He combines deep analytic problem-solving from research—where his work on GR and QFT drew popular-press attention—with hands-on contributions to open-source tooling (notably fixes and features in the Determined platform and a Stable Diffusion example). Garrett routinely leads novel research lines, mentors junior researchers, and manages complex project logistics, making him comfortable both writing papers and shipping robust ML systems. An interdisciplinary collaborator across 15+ international co-authors, he brings a rare mix of theoretical rigor and pragmatic engineering focused on scalable, reproducible AI.
7 years of coding experience
10 years of employment as a software developer
Bachelor of Science - BS, Bachelor of Science - BS at University of MarylandPhD, Theoretical Physics, PhD, Theoretical Physics at University of Pennsylvania
Github Skills (11)
pytorch10
machine-learning10
distributed-training10
deep-learning10
python10
mlops9
data-science8
tensorflow7
hyperparameter-optimization7
kubernetes-pods5
kubernetes5
Programming languages (9)
Github contributions (5)
determined-ai/determined
Apr 2022 - Jan 2023
Determined is an open-source machine learning platform that simplifies distributed training, hyperparameter tuning, experiment tracking, and resource management. Works with PyTorch and TensorFlow.
Role in this project:
ML Engineer
ML EngineerContributions:321 reviews, 12 commits, 138 PRs in 8 months
Contributions summary:Garrett contributed to the Determined AI platform by implementing and fixing issues related to PyTorch trials, including synchronizing epoch indices across workers and introducing new PyTorch callbacks. They addressed bugs and enhanced existing functionality related to distributed training. The user also worked on integrating new features, such as a Stable Diffusion textual inversion example.
pytorchdata-sciencedeep-learningdistributed-trainingmachine-learning
garrett361/mamba
Dec 2024 - Apr 2025
Mamba SSM architecture
Contributions:5 PRs, 147 pushes, 12 branches in 3 months
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.