About the job
In the DeepMind's Frontier Health team, you will build agentic, multimodal AI systems that meaningfully assist in healthcare. You will develop novel methods at the intersection of large language models, reinforcement learning, and clinical medicine, contributing to publications while shaping real-world products. As a passionate researcher, you will grow in a fast-moving, collaborative environment spanning machine learning, clinical expertise, and product development.
Responsibilities
Develop reinforcement learning pipelines (e.g., RLHF, GRPO, reward modeling) to align AI systems with strict clinical quality and safety standards.
Build evaluation frameworks, and human-expert grading protocols to benchmark AI on realistic clinical scenarios.
Identify key problems, formulate hypotheses, run scalable experiments, and publish findings at venues.
Partner closely with clinicians, product teams, and ML researchers to translate theoretical research into tangible healthcare impact.
Design and evaluate novel techniques for medical AI, focusing on agentic clinical reasoning, structured tool use, and multimodal understanding (text, image, audio/video).
Qualifications
Minimum
PhD degree in Computer Science, a related field, or equivalent practical experience.
2 years of experience leading a research agenda.
One or more scientific publication submissions for conferences, journals, or public repositories (such as CVPR, ICCV, NeurIPS, ICML, ICLR, etc.).
Preferred
Experience collaborating with domain experts or clinicians, and interest in responsible AI (e.g., safety).