Yotam Doron
Senior Staff Research Engineer at Google DeepMind
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolYotam Doron is a Senior Staff Research Engineer based in London with 15 years of experience building and maintaining research-grade software for reinforcement learning and physics-based simulation. At DeepMind/Google DeepMind he progressed through research-engineering roles to senior staff, shipping improvements to core RL tooling and modernizing environment interfaces used by the community. He contributes to high-profile open-source projects such as DeepMind’s bsuite and dm_control, focusing on code quality, refactoring and usability enhancements that keep research stacks robust and maintainable. Comfortable bridging research and production, he combines deep engineering discipline with practical ML environment design. His background includes an EngD research trajectory and a stint as a startup co-founder, underscoring both academic rigor and product-minded initiative. Colleagues rely on him for steady, low-noise improvements that reduce technical debt and accelerate experimentation.
15 years of coding experience
13 years of employment as a software developer
BA, BA at University of CambridgeMSc, MSc at University College London, U. of London
Github Skills (14)
machine-learning10
python10
reinforcement-learning10
tensorflow9
deeplearning-ai9
deep-learning9
refactoring8
simulation8
simulations8
neural-network8
artificial-neural-networks8
mujoco7
api5
apim5
Programming languages (2)
Github contributions (5)
google-deepmind/bsuite
Aug 2019 - Apr 2020
bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent
Role in this project:
Backend Developer
Backend DeveloperContributions:7 reviews, 9 commits, 3 PRs in 8 months
Contributions summary:Yotam primarily focused on code maintenance and updates within the `bsuite` repository. Their contributions included removing unused imports, updating GymWrapper to GymFromDMEnv names, using six.add_metaclass, and updating nest imports. These changes suggest involvement in code cleanup and refactoring, aimed at improving code quality and aligning with best practices. The commits indicate a focus on maintaining and improving the codebase's structure and dependencies.
stable-baselinesagentreinforcement-learningdeep-reinforcement-learningcapabilities
google-deepmind/dm_control
Jan 2018 - Sep 2019
Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
Role in this project:ML Engineer
ML EngineerContributions:11 commits, 1 PR, 4 pushes in 1 year 8 months
Contributions summary:Yotam primarily contributed to the dm_control library by adding environment-related features and improving existing functionalities. They added environment keyword arguments and expanded the actuator-action correspondence within the suite, improving usability and clarity. Additionally, the user updated documentation and replaced older environment interface code with `dm_env` imports, modernizing the codebase and ensuring compatibility.
simulationreinforcementsoftware-stackphysics-based-simulationsimulation-based
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.