Research Scientist, Science AI, Google Research

Google•Cambridge, MA

1d•$147,000 - $210,000

About The Position

Google Research is seeking a Research Scientist to develop agentic AI systems for scientific discovery. The role focuses on integrating domain-specific tools, simulators, datasets, and models—such as genomics pipelines, structure prediction, and physics-based simulation—so agents can use them reliably at scale. The scientist will collaborate with researchers inside and outside Google to identify high-impact problems, validate agent-generated hypotheses, and connect prediction with discovery. The position includes setting and leading a research agenda at the intersection of agents and science, publishing research, and helping shape Google's long-term strategy for AI-accelerated science. It also involves owning robust, scalable infrastructure for agent execution, orchestration, and evaluation so research ideas can become reliable systems others can build on. Google Research conducts fundamental research and contributes to product development and infrastructure. Research Scientists design experiments, prototype implementations, develop architectures, and share findings with the wider research community through publications and other forms of research sharing.

Requirements

PhD in Computational Biology, Bioinformatics, Biophysics, Physics, Computer Science, or a related quantitative field, or equivalent practical experience.
Experience developing autonomous agents, multi-agent workflows, or tool-using LLM systems.
One or more first-author scientific publications in peer-reviewed journals or conferences.

Nice To Haves

Two years of experience leading a research agenda in computational biology, computational physics, or machine learning for science.
Experience developing and deploying machine learning models or LLM-based systems using modern frameworks such as JAX, PyTorch, or TensorFlow.
Coding experience in Python, C++, or a similar language.
Experience designing benchmarks, evaluations, or verifiable reward signals for scientific reasoning, data analysis, or code generation.

Responsibilities

Design, build, and evaluate agentic AI systems for scientific discovery.
Integrate domain-specific tools, simulators, datasets, and models so agents can use them reliably at scale.
Collaborate with scientists inside and outside Google to identify high-impact problems, validate agent-generated hypotheses, and connect prediction with discovery.
Set and lead a research agenda at the intersection of agents and science; identify new directions, publish in top ML venues, and help shape long-term strategy for AI-accelerated science.
Own robust, scalable infrastructure for agent execution, orchestration, and evaluation, and help turn research ideas into reliable systems others can build on.
Set up large-scale tests and deploy promising ideas quickly and broadly while managing deadlines and deliverables.
Contribute to the wider research community by sharing and publishing findings.