Research Scientist, Robotics Pre-Training and Data Quality, DeepMind

• Drive data quality, acquisition, and labeling strategies, and oversee mixture optimization for robotics foundation models.
• Coordinate training runs, improve training recipes, and systematically hill-climb model performance against benchmarks.
• Build and maintain the software tools and processes required to iterate on research ideas quickly and verify data quality through robot policy training.
• Contribute to our wider research agenda, including reinforcement learning, vision-language-action modeling, world-action models, and simulation, working in a fast-paced, collaborative team environment.

Minimum qualifications:

• PhD in a technical field or equivalent practical experience.
• Experience with algorithmic architectures, data sources, and training and inference techniques for generative multimodal models, especially for robotics applications (e.g., VLAs, WAMs).
• Experience optimizing models and systems (e.g., performance tuning, experimentation, and debugging).

Preferred qualifications:

• Practical experience in "hill-climbing" model performance (e.g., iteratively improving training recipes and model benchmarks).
• Experience working with simulators and real-world robotic platforms (e.g., dexterous manipulation, multimodal sensing).
• Experience with benchmarking, different data sources, data labeling strategies, and data mixture optimization for robotics foundation models.
• A passion for bringing research from the lab to robust, real-world robotic systems.
• Proven ability to build and maintain the software tools for rapid iteration on research ideas.

• PhD in a technical field or equivalent practical experience.
• Experience with algorithmic architectures, data sources, and training and inference techniques for generative multimodal models, especially for robotics applications (e.g., VLAs, WAMs).
• Experience optimizing models and systems (e.g., performance tuning, experimentation, and debugging).