Tom Hennigan
Senior Staff Software Engineer at Google DeepMind
England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolTom Hennigan is a Senior Staff Software Engineer with 12 years of experience building compute infrastructure for large-scale AI research, currently leading work on Gemini at Google DeepMind. He combines deep systems and ML engineering expertise—spanning XLA, JAX and TensorFlow—with hands-on contributions to compiler and distributed runtime features that improve robustness and usability for GPU/TPU workloads. Tom has led teams that transitioned DeepMind from TensorFlow to JAX, authored major LLM publications (including Gopher and Chinchilla) and developed core libraries like Haiku and Sonnet used across the org. His background ranges from building tera-scale data pipelines and spot-instance orchestration at DueDil to shipping production Google products, reflecting a blend of research-grade systems design and pragmatic production engineering. Notably, he contributes to high-profile open-source projects (XLA, JAX, TensorFlow) where he focuses on Python client usability, distributed primitives and testing/build infrastructure.
12 years of coding experience
12 years of employment as a software developer
Dr Challoners Grammar SchoolMEng, Computer Science, First Class Honours, MEng, Computer Science, First Class Honours at University of Southampton
English, French
Stackoverflow
1,092reputation
123kreached
9answers
1question
Github Skills (33)
probabilistic-programming10
python10
testing10
machine-learning10
build-system10
cicd10
deeplearning-ai10
deep-learning10
tensorflow10
parallel-computing10
eager-execution10
neural-network10
xla10
jax10
c-language9
Programming languages (7)
Github contributions (5)
google-deepmind/sonnet
Jun 2018 - Dec 2022
TensorFlow-based neural network library
Role in this project:
ML Engineer
ML EngineerContributions:4 releases, 3 reviews, 202 commits in 4 years 7 months
Contributions summary:Tom made several contributions focused on testing and integrating eager mode compatibility within the TensorFlow-based neural network library, Sonnet. Their work involved adding and updating tests for various modules, particularly related to convolutional neural networks and AlexNet, to ensure proper functionality in eager execution. They further addressed issues with Autograph and the creation of variables within the library, along with ensuring proper execution of these modules.
deep-learningneural-networksmachine-learningneural-networktensorflow
google-deepmind/dm-haiku
Jan 2020 - Jan 2023
JAX-based neural network library
Role in this project:Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:16 releases, 33 reviews, 328 commits in 3 years
Contributions summary:Tom's contributions focused on improving the testing and build infrastructure of the Haiku library. This included implementing PyTest for running tests within the CI environment, which resolved issues with Bazel and Python's path handling. The user added support for parallel test execution, and also made changes to the test configuration. Furthermore, the user was responsible for improving the build system for documentation and general testing in this repository.
deep-learningneural-networksmachine-learningneural-networkdeep-neural-networks
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.