Jake Mathern
Senior SWE at Microsoft
Redmond, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolJake Mathern is a Senior Software Engineer in Redmond with 10 years of experience focused on high-performance graphics, gaming, and machine learning infrastructure. A DigiPen alum and long-time Microsoft engineer, he currently works on DirectML and has contributed substantive C++ fixes and opset support to the widely used ONNX Runtime project. He prefers work that rewards low-level optimization—cache-coherent access patterns, tight memory and bandwidth constraints—and brings practical game-dev instincts to systems and AI engineering. Known for improving operator implementations across execution providers, he blends hands-on debugging with cross-team integration to ship robust, high-throughput components.
10 years of coding experience
Real Time Interactive Simulation C++ Vulkan Team projects, Real Time Interactive Simulation C++ Vulkan Team projects at DigiPen Institute of TechnologyMathematics and Computer Science, Mathematics and Computer Science at Normandale Community College
English
Github Skills (11)
neural-network10
machine-learning10
c-language10
deep-learning10
onnx10
cprogramming-language10
hardware-acceleration10
tensorflow9
ai-framework9
pytorch9
cmake7
Programming languages (4)
Github contributions (5)
microsoft/onnxruntime
May 2020 - Nov 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:
ML Engineer
ML EngineerContributions:8 reviews, 12 commits, 21 PRs in 2 years 6 months
Contributions summary:Jake primarily contributed to the ONNX Runtime project by fixing issues related to operator implementations and improving the integration of new ONNX operators. Their work involved modifying C++ code within the DirectML and CPU execution providers to address bugs in the Concat operator, enable tests for the Pad operator, and add support for new ONNX opset 12 operators (Clip, Max, Min, MaxPool, ReduceMax, ReduceMin, Celu, GreaterOrEqual, LessOrEqual, ArgMin/Argmax). These changes included updates to operator registration and schema definitions, enhancing the functionality of the runtime environment. Furthermore, the user modified the code to refine the behavior of the CPU EP when ConvTranspose output shape is specified.
runtimetrainingtensorflowai-frameworkaccelerator
Jamather/onnxruntime
Jun 2022 - Sep 2024
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Contributions:1 PR, 22 pushes, 11 branches in 2 years 3 months
pytorchdeep-learningruntimemachine-learningonnx
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.