James Allingham
Research Scientist at Google DeepMind
London, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolJames Allingham is a research scientist at Google DeepMind with a decade of machine learning experience and a PhD from the University of Cambridge, where he focused on Bayesian deep learning and deep generative models. His work blends rigorous probabilistic foundations with practical engineering—spanning research internships at Google Brain, a visiting stint at the University of Amsterdam, and production-focused contributions to Wolfram's deep learning framework. He has co-authored papers on automated prompt engineering and interactions between ensembles and sparse Mixture-of-Experts models, and has contributed to the ONNX project by improving operator test coverage and documentation. Based in London, he pairs academic depth with hands-on implementation skills across PyTorch/TensorFlow and production tooling, and he brings an uncommon mix of open-source stewardship and applied research that helps move models from prototype to robust deployments.
10 years of coding experience
2 years of employment as a software developer
Bachelor of Science in Engineering Electrical and Information Engineering, Bachelor of Science in Engineering Electrical and Information Engineering at University of the WitwatersrandDoctor of Philosophy Information Engineering, Doctor of Philosophy Information Engineering at University of CambridgeParktown Boys' High School
Afrikaans, English
Stackoverflow
716reputation
43kreached
3answers
22questions
Github Skills (16)
machine-learning10
deep-learning10
onnx10
python10
testing10
numpy9
deep-neural-networks9
pytorch8
tensorflow8
ubuntu6
neo4j6
cmu-sphinx6
css6
nodejs6
azure6
Programming languages (13)
Github contributions (5)
onnx/onnx
Sep 2019 - Sep 2019
Open standard for machine learning interoperability
Role in this project:
ML Engineer
ML EngineerContributions:5 commits, 10 PRs, 35 comments in 4 days
Contributions summary:James contributed significantly to the testing framework within the ONNX repository, specifically focusing on the `Unsqueeze` and `Gemm` operators. Their work involved adding new test cases to cover various scenarios, including tests for different axes, negative axes, and cases where the bias is optional. The user also updated the operator documentation and test coverage. These changes demonstrate a focus on improving the quality and robustness of the ONNX operator library.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
JamesAllingham/AutoImpute
Feb 2018 - Aug 2024
Contributions:2 PRs, 105 pushes, 3 branches in 6 years 7 months
acsimputation
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.