Contribute to the creation of novel multi-agent learning algorithms and frameworks, with a focus on performance and scalability in JAX. Build and maintain simulation platforms and end-to-end research pipelines to run experiments on DeepMind’s cutting-edge infrastructure, including massive TPU pods. Partner with Research Scientists to transform mathematical concepts and research hypotheses into robust, production-quality code and reproducible experiments.
Lead the engineering direction for complex research projects, establishing best practices for code quality and maintainability, and mentoring junior engineers on the team. Optimize every part of the research workflow, from data processing and model training to results analysis, to accelerate the pace of discovery.
Job description provided by Google.
View the original posting
