Marie White
Staff Software Engineer at Google
Switzerland
Summary
🤩
RockstarMarie White is a Staff Software Engineer based in Sydney with eight years of professional experience and a decade-plus background in software and electronics roles spanning startups and tech giants. At Google she focuses on ML compiler and runtime optimization, contributing to high-impact open-source projects like IREE and SHARK where she integrated, benchmarked and tuned models such as MobileBERT, MobileNetV3, ResNet50, BERT and GPT-2 across hardware targets. She combines systems-level engineering (XLA, TF device configs, TF32, torch-inductor) with practical benchmarking and tooling to move research models into high-performance production paths. Known for bridging compiler internals and end-to-end ML workflow automation, she brings both hands-on implementation and an eye for reproducible performance comparisons across runtimes.
9 years of coding experience
8 years of employment as a software developer
Stackoverflow
36reputation
505reached
1answer
0questions
Github Skills (23)
benchmark10
performance-monitor10
performance-analytics10
python10
machine-learning10
benchmarking10
performance-measurement10
deep-learning10
performance-analysis10
tensorflow10
performance-tuning10
tflite10
performance-monitoring10
machine-learning-models9
pytorch9
Programming languages (4)
Github contributions (5)
iree-org/iree
Mar 2022 - Jan 2023
A retargetable MLIR-based machine learning compiler and runtime toolkit.
Role in this project:
ML Engineer
ML EngineerContributions:347 reviews, 36 commits, 223 PRs in 10 months
Contributions summary:Marie contributed code related to integrating and benchmarking machine learning models within the IREE compiler and runtime toolkit. They implemented and tested new quantized machine learning models, specifically Mobilebert, Mobilenet V3, and Resnet 50, adding them to presubmit tests. Additionally, they wrote scripts to compare performance between TFLite and IREE, covering various hardware and software configurations. Finally, the user created and updated tools for generating and displaying benchmark reports.
compilermachine-learningmlirvulkantensorflow
nod-ai/AMD-SHARK-Studio
Nov 2022 - Jan 2023
AMD-SHARK Studio -- Web UI for SHARK+IREE High Performance Machine Learning Distribution
Role in this project:ML Engineer
ML EngineerContributions:7 commits, 33 PRs, 12 pushes in 1 month
Contributions summary:Marie primarily focused on enabling and optimizing TensorFlow (TF) models within the SHARK (SHARK+IREE) framework. They modified code to integrate the XLA compiler for TF models, improving performance. Their contributions extended to various TF models, including BERT, MiniLM, GPT2, and others. They updated the code to use the correct TF device configuration and added support for TF32 and torch-inductor.
amdmachine-learningdeep-learningmlirpytorch
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.