Lead ML Engineer — Foundation Models for Drug Discovery

Roche; South San Francisco (CA); On-site; USD 193,000 - 358,000; Full time

Job summary

Roche’s AI for Drug Discovery group seeks a Principal Machine Learning Engineer to advance autonomous agents and foundation models in drug discovery. You will work across engineering and research, building scalable ML systems and robust production pipelines, partnering with Genentech scientists to translate complex problems into efficient, end-to-end workflows.

You will lead the MLOps/AgentOps stack, define the engineering roadmap, and guide cross-functional platforms while applying frontier AI

Qualifications

• PhD/MS/BS with extensive ML infra leadership and track record.
• Strong Python programming and software engineering fundamentals.
• Hands-on experience with PyTorch/JAX and distributed training on AWS or HPC.
• Experience designing agent orchestration frameworks and persistent memory.
• Interest in AI for Drug Discovery and biology/chemistry.

Responsibilities

• Architect and deploy autonomous agents for drug discovery workflows.
• Design scalable ML systems and productionize foundation models.
• Own MLOps/AgentOps stack: tracking, observability, evaluation, CI/CD.
• Define long-term engineering roadmap for AI4DD; collaborate with Genentech leadership.
• Translate scientific problems into scalable, efficient systems.

Skills

Python programming

Git

CI/CD

Software engineering fundamentals

Distributed systems

Education

PhD in CS/ML/Engineering

MS in CS/ML/Engineering

BS in CS/Eng

Tools

LangGraph

MCP-based tool integration

PyTorch

JAX

AWS

HPC environments

Job description

Roche’s AI for Drug Discovery group seeks a Principal Machine Learning Engineer to advance autonomous agents and foundation models in drug discovery. You will work across engineering and research, building scalable ML systems and robust production pipelines, partnering with Genentech scientists to translate complex problems into efficient, end-to-end workflows.

You will lead the MLOps/AgentOps stack, define the engineering roadmap, and guide cross-functional platforms while applying frontier AI