Research Engineer, Training Infrastructure
Description For Research Engineer, Training Infrastructure
At Google DeepMind, we're at the forefront of artificial intelligence research and development. Our training infrastructure team in Robotics is seeking a Research Engineer to play a crucial role in realizing general intelligence on physical robotic platforms. You'll be part of a diverse team working on cutting-edge robotics research, developing and maintaining robust and scalable training pipelines that transform large-scale robotics data into high-quality perception and control agents.
In this role, you'll collaborate closely with research scientists, engineers, and stakeholders across DeepMind. Your responsibilities will span the entire training pipeline, from data acquisition and augmentation to model training and evaluation. You'll be involved in designing and debugging new model architectures, conducting experiments to optimize performance, deploying agents on real robots, and building new robotics datasets.
We're looking for candidates with a strong background in computer science, mathematics, or related fields, with solid Python skills and a deep understanding of machine learning and statistics. Experience with robotics systems, ML frameworks like JAX, and handling large-scale datasets would be highly advantageous.
At Google DeepMind, we value diversity and are committed to equal employment opportunity. We welcome candidates from all backgrounds and experiences, and we're dedicated to creating an inclusive environment where everyone can thrive and contribute to our mission of advancing AI for the benefit of humanity.
Join us in shaping the future of robotics and AI, where your work will have a direct impact on pushing the boundaries of what's possible in machine learning and robotics. Be part of a team that's not just imagining the future, but actively building it.
Responsibilities For Research Engineer, Training Infrastructure
- Collaborate with colleagues to design, build, and debug new model and agent architectures
- Conduct experiments to analyse the effect of different architecture choices, hyperparameters, and dataset weights on model performance
- Deploy and evaluate trained agents on real robots
- Build and curate new robotics datasets for training, ensuring data quality and diversity
- Investigate and diagnose performance regressions in the training pipeline, identifying and resolving bottlenecks
- Balance long-term engineering goals with the need to support short-term research experimentation
Requirements For Research Engineer, Training Infrastructure
- BSc, MSc, or PhD/DPhil degree in computer science, mathematics, applied statistics, machine learning, or a related field, or equivalent industry experience
- Proven knowledge of Python
- Solid understanding of machine learning and statistics
Benefits For Research Engineer, Training Infrastructure
- Equal employment opportunity
