Scott Mckay
Principal Software Development Engineer at Microsoft
Gold Coast City, Queensland, Australia
Summary
🤩
RockstarScott Mckay is a Principal Software Development Engineer based on the Gold Coast with a decade of experience building high-performance, highly reliable global services at Microsoft. He combines deep expertise in C# and C++ with pragmatic object-oriented design to architect, implement and support low-latency, production-grade systems—most recently contributing to the ONNX Runtime inference framework and improving operator performance and stability. A proven incident resolver and runtime debugger, Scott thrives on diagnosing critical production issues in real time and driving measurable performance optimizations. He has led multiple technical teams and platforms at Microsoft—from query fabrics and social graph services to privacy-aware personal data APIs—championing code quality, testing, and maintainability. Beyond shipping systems, he contributes to open-source ML infrastructure work on ONNX, improving specification clarity and type/shape inferencing for core operators. Fast-learning and exceptionally fast at turning ideas into robust code, he blends hands-on engineering with mentorship and architectural leadership.
10 years of coding experience
Stackoverflow
323reputation
31kreached
15answers
2questions
Github Skills (21)
c-language10
machine-learning10
inference10
onnx10
deep-learning10
onnxruntime10
cuda10
cprogramming-language10
shapes10
tensorrt9
performance-optimization9
neural-network9
numerical-computing9
documentation9
python8
Programming languages (9)
Github contributions (5)
microsoft/onnxruntime
Nov 2018 - Jan 2023
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:
ML Engineer
ML EngineerContributions:3516 reviews, 711 commits, 1112 PRs in 4 years 2 months
Contributions summary:Scott's commits primarily involve implementing and optimizing functionalities within the ONNX Runtime, a cross-platform machine learning inferencing and training accelerator. Their contributions focus on enhancing the performance of the `Slice` and other operators, which involved modifying kernel implementations and addressing potential numerical instability issues with bfloat16 kernels. Furthermore, they also contributed to the model building and validation framework through the addition of unit tests and validation of key core functionalities.
runtimetrainingtensorflowai-frameworkaccelerator
onnx/onnx
Apr 2018 - Mar 2020
Open standard for machine learning interoperability
Role in this project:ML Engineer
ML EngineerContributions:7 reviews, 19 commits, 18 PRs in 1 year 11 months
Contributions summary:Scott focused on improving the documentation and clarifying the specifications of operators within the ONNX framework. Their work included refining descriptions of the `FeatureVectorizer` and `Random*Like` operators. Additionally, the user addressed issues related to RNN/LSTM/GRU recurrent weights, enhancing consistency. Further contributions involved enabling type/shape inferencing for operators such as If and Loop, and fixing a typo in Scan's shape inferencing.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
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.