ML Scientist: Autonomous Agents for Small Molecule Design

Roche; San Francisco (CA); On-site; USD 148,000 - 274,000; Full time

Benefits offered by this job

Discretionary annual bonus

Job summary

Roche Genentech is seeking an experienced ML Scientist to create autonomous, LLM-driven agentic workflows for small-molecule drug design. You will orchestrate ML models, physics-based methods, and cheminformatics tools with chemists and structural biologists to accelerate discovery.

The role requires strong ML fundamentals, Python, and hands-on experience with PyTorch/JAX, RDKit/OpenEye, and LangChain in a collaborative research environment.

Qualifications

• Experience building LLM-driven agents for scientific workflows.
• Strong foundations in linear algebra, probability, optimization.
• Hands-on with PyTorch or JAX and modern agentic coding environments.
• PhD or equivalent in ML/CS/chemical engineering or related field.

Responsibilities

• Design and implement agentic workflows for small-molecule drug design.
• Fine-tune foundation models using internal/external datasets and tools.
• Collaborate with chemists and structural biologists to validate hypotheses.
• Publish results, contribute to open-source releases and talks.
• Collaborate across Roche Genentech computational teams and partners.

Skills

Python

LangChain

PyTorch

JAX

Reinforcement Learning

GNNs

Sequence/Language Models

Cheminformatics tools

Orchestration of tools/models

Education

PhD in ML/CS/Chemical Eng/Physics

Tools

RDKit

OpenEye

Job description

Roche Genentech is seeking an experienced ML Scientist to create autonomous, LLM-driven agentic workflows for small-molecule drug design. You will orchestrate ML models, physics-based methods, and cheminformatics tools with chemists and structural biologists to accelerate discovery.

The role requires strong ML fundamentals, Python, and hands-on experience with PyTorch/JAX, RDKit/OpenEye, and LangChain in a collaborative research environment.