Brian Martin
Principal Software Engineer Team Lead at Microsoft
Redmond, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolBrian Martin is a Principal Software Engineer and team lead at Microsoft with over eight years in senior engineering roles and a two-decade career contributing to Windows, Office, Surface Hub, HoloLens, and Windows Mixed Reality. He builds developer-facing ML inference APIs and tooling for Windows, created automated test suites and release pipelines, and partners with third-party developers to integrate on-platform AI. His open-source contributions include performance-focused optimizations to ONNX Runtime’s quantization operators and practical samples for Windows Machine Learning, showing a blend of low-level performance work and developer experience improvements. With a B.Sc. in Physics from UBC, he pairs analytical rigor with hands-on systems engineering and a knack for turning device- and OS-level complexity into reliable developer tools.
7 years of coding experience
Bachelor of Science (B.Sc.) Honours, Physics, Bachelor of Science (B.Sc.) Honours, Physics at The University of British Columbia
Stackoverflow
80reputation
1kreached
9answers
0questions
Github Skills (23)
c-language10
windows10
machine-learning10
onnx10
deeplearning-ai10
deep-learning10
cprogramming-language10
hardware-acceleration9
tensorflow9
neural-network8
testing8
artificial-neural-networks8
ai7
arm6
hololens6
Programming languages (4)
Github contributions (5)
microsoft/Windows-Machine-Learning
Jun 2019 - Jan 2023
Samples and Tools for Windows ML.
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 88 reviews, 37 commits in 3 years 8 months
Contributions summary:Brian primarily contributed to the Windows Machine Learning repository by implementing samples and making improvements to existing ones. They integrated a model from an embedded resource, demonstrating knowledge of model loading and integration. Furthermore, the user addressed layout and resizing issues within the sample gallery, added tooltips, and improved the clarity of the code.
pytorchcaffe2pythonwindowswinmltools
microsoft/onnxruntime
Dec 2018 - Sep 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:ML Engineer
ML EngineerContributions:169 reviews, 45 commits, 58 PRs in 3 years 10 months
Contributions summary:Brian focused on optimizing and testing the `DequantizeLinear` and `QuantizeLinear` operators within the ONNX Runtime, improving performance for quantized machine learning models. The commits include code modifications to the test suite, adding and modifying tests for different scenarios, and changes to the CPU implementation of these operators. They also addressed warnings and incorporated telemetry for event timing and measurements related to tensor processing, showing an emphasis on performance and monitoring.
runtimetrainingtensorflowai-frameworkaccelerator
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.