Research Scientist, Multisensor Robotics, DeepMind

Requirements:

• PhD degree in a relevant quantitative field (e.g., Robotics, Computer Science, Machine Learning), or equivalent practical experience.
• 5 years of professional research experience in an industrial research and development (R&D) laboratory.
• 3 years of experience designing, training, and scaling reinforcement learning (RL), imitation learning (IL), or multimodal generative architectures (including Vision-Language-Action (VLA) or video foundation models).
• 2 years of experience optimizing foundation models for production, utilizing techniques like distillation, model pruning, and post-training quantization to meet on-device latency and memory footprint constraints.

Nice to haves:

• 3 years of experience building data pipelines, deploying production-ready systems, and experience managing multi-sensory datasets.

What you'll be doing:

• Architect and train models to expand robotic capabilities and prototype concepts.
• Apply domain expertise in sim-to-real transfer, imitation learning, Vision-Language-Action (VLA) models, multi-sensory learning (tactile, audio), and model optimization (distillation, pruning, quantization).
• Propose novel ideas, address manipulation hardware requirements, and communicate results to audiences.

Perks and benefits:

• Google maintains a portfolio of research projects driven by fundamental research, new product innovation, product contribution and infrastructure goals.
• Freedom to emphasize specific types of work.
• Opportunity to work on real-world problems that span the breadth of computer science.