Andrew Trask
Executive Director at University of Oxford
Oxford, England, United Kingdom
Summary
🤩
RockstarAndrew Trask is a leader in privacy-preserving machine learning and deep learning with 12 years of experience spanning research, open-source stewardship, and education. He leads the OpenMined community, is a Senior Research Scientist at DeepMind, and pursues a PhD at Oxford, blending academic rigor with large-scale industry research. Author of Grokking Deep Learning and contributor to its companion repository, he translates complex ideas into teachable code and curricula. His open-source work on PySyft and contributions to Udacity’s Private AI course highlight a practical focus on differential privacy and secure ML techniques such as PATE. Based in Oxford, he also serves in policy-adjacent roles like a CFR Term Member, reflecting interest in the societal impact of AI. Colleagues know him for combining hands-on model development with community-building that makes privacy tools accessible.
13 years of coding experience
Github Skills (13)
web-framework10
pytorch10
machine-learning10
deep-learning10
differential-privacy10
python10
data-science10
numpy10
sym10
tensorflow9
mnist9
keras9
notebook8
Programming languages (12)
Github contributions (5)
iamtrask/Grokking-Deep-Learning
Aug 2016 - Feb 2020
this repository accompanies the book "Grokking Deep Learning"
Role in this project:
Data Scientist
Data ScientistContributions:25 commits, 6 PRs, 27 pushes in 3 years 6 months
Contributions summary:Andrew primarily worked on implementing and refining deep learning models within the "Grokking Deep Learning" repository, focusing on chapter-specific code. Their contributions include code implementations related to chapter 10 and 13, along with modifications in chapter 14, showcasing a progression of learning and model refinement. The user appears to be actively engaged in exploring and implementing deep learning concepts and applying them to the MNIST dataset.
deep-learning
OpenMined/PySyft
Jul 2017 - Dec 2022
Perform data science on data that remains in someone else's server
Role in this project:Data Scientist
Data ScientistContributions:6 releases, 302 reviews, 4130 commits in 5 years 5 months
Contributions summary:Andrew's contributions center around enhancing the SimpleService functionality within the PySyft project. Their work involved empowering simple-service messages to carry arbitrary payloads, improving the simple service's handling of messages, and fixing a bug related to database table creation for hagrid-launched Domains. This suggests a focus on extending the system's capabilities and fixing core functionality related to a specific, key service in the context of data science.
data-sciencedeep-learningsecure-computationpytorchprivacy
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.