Staff ML Research Scientist: Next-Gen Training Architect

DeepMind Technologies Limited; Mountain View (CA); On-site; USD 207,000 - 300,000; Full time

Job summary

DeepMind Technologies Limited in Mountain View, CA, seeks a senior researcher to invent ML architectures, optimizers, and training methods, and to build the systems that let Gemini propose, implement, and test such changes on its own. You will be the team's final technical verifier, deciding which AI-generated ideas are real advances and carrying the winners into next-generation Gemini models.

This is a deep individual-contributor role.

Qualifications

• PhD in Computer Science, a related field, or equivalent practical experience.
• 4 years of experience with research agendas across multiple teams or projects.
• One or more scientific publication submission(s) for conferences, journals, or public repositories.
• Experience working with deep learning frameworks (e.g., JAX/Flax, PyTorch) for distributed systems and accelerator clusters.

Responsibilities

• Invent, implement, and scale new ML algorithms, architectures, optimizers, training objectives, and recipes.
• Design and build automated research systems enabling Gemini to propose, code, and evaluate changes to its training stack.
• Serve as final verifier for AI-proposed algorithm changes: reproduce results and rule out confounds.
• Build and own lean, high-performance training and evaluation code (JAX on TPUs, with custom kernels).
• Partner with Gemini teams to land improvements in production models and share results.

Skills

Deep learning

Research experience

Publications

Distributed systems

Education

PhD in CS or related field

Equivalent practical experience

Tools

JAX

Flax

PyTorch

Job description

DeepMind Technologies Limited in Mountain View, CA, seeks a senior researcher to invent ML architectures, optimizers, and training methods, and to build the systems that let Gemini propose, implement, and test such changes on its own. You will be the team's final technical verifier, deciding which AI-generated ideas are real advances and carrying the winners into next-generation Gemini models.

This is a deep individual-contributor role.