Research Scientist, Agent Post-Training, DeepMind

Requirements:

• Bachelor's or Master’s degree in Computer Science, Machine Learning, a related quantitative field, or equivalent practical experience.
• 2 years of experience with reinforcement learning (RL), supervised fine-tuning (SFT), or agent-based systems, including modeling or applied infrastructure.
• 2 years of experience with machine learning frameworks such as JAX, Flax, TensorFlow, or PyTorch, and scaling experiments.

Nice to haves:

• PhD in Computer Science, Machine Learning, or a related quantitative field.
• Publications in reinforcement learning (RL), tool-use, or agentic systems, or applied projects demonstrating infrastructure for agentic setups.
• Experience balancing research velocity and code quality, including managing technical debt.
• Ability to investigate complex technical issues and develop high-standard scientific artifacts.
• Preference for simple, elegant solutions to research problems.
• Experience with systems design and managing code complexity in large codebases.

What you'll be doing:

• Lead the full research process, from forming hypotheses to delivering direct impact in Gemini post-training.
• Design effective ablations to systematically validate and iterate on research ideas.
• Communicate research findings and experimental results through clear plots and reports.
• Develop reliable infrastructure and utilities to support post-training research and evaluations.
• Collaborate with other researchers and engineers to accelerate project velocity.

Perks and Benefits:

• Work in a unique culture and environment where long-term ambitious research can flourish.
• Opportunity to drive research that makes up the backbone of upcoming releases.
• Create extraordinary impact with various experiences and backgrounds valued.
• Global teams offer various learning opportunities and varied career pathways for achieving exceptional results.