Ted Themistokleous
Software Engineer at AMD
Burlington, Ontario, Canada
Summary
🤩
RockstarTed Themistokleous is a software engineer based in Burlington, Ontario with four years of hands-on experience specializing in ML inference and performance optimization. He contributes to high-profile open-source work on microsoft/onnxruntime, improving the MIGraphX execution provider with features like stream synchronization, int8 quantization, and broader operator support useful for models such as stable diffusion. Ted combines practical engineering—fixing benchmarking integration and legacy Python compatibility—with a focus on reducing runtime overhead for production ML workloads. Outside code, he's a guitar-playing programmer, suggesting a creative, patient approach to problem solving.
3 years of coding experience
Github Skills (11)
machine-learning10
c-language10
onnx10
cprogramming-language10
hardware-acceleration10
roc9
python8
deeplearning-ai7
deep-learning7
tensorflow5
pytorch5
Programming languages (5)
Github contributions (5)
microsoft/onnxruntime
Oct 2022 - Nov 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:
ML Engineer
ML EngineerContributions:71 reviews, 3 commits, 58 PRs in 1 month
Contributions summary:Ted primarily contributes to the MIGraphX execution provider for ONNX Runtime, focusing on enhancing its capabilities for machine learning inference. Their work involves implementing stream synchronization for performance optimization within the MIGraphX provider, leading to reduced overhead. They also address integration issues with the `eval_squad.py` script and other benchmarking tools, adding support for multiple execution providers and older Python versions. Furthermore, they add support for various ONNX operators, including those needed for stable diffusion models, as well as features such as int8 quantization and model saving/loading.
runtimetrainingtensorflowai-frameworkaccelerator
TedThemistokleous/onnxruntime
Sep 2022 - Jan 2025
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Contributions:159 pushes, 60 branches in 2 years 5 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.