Machine Learning Scientist - Agents for Applied Small Molecule Drug Design
Roche, Danesbury, Welwyn Hatfield
Machine Learning Scientist - Agents for Applied Small Molecule Drug Design
Salary not available. View on company website.
Roche, Danesbury, Welwyn Hatfield
- Full time
- Permanent
- Onsite working
Posted 1 day ago, 21 Jun | Get your application in today.
Closing date: Closing date not specified
Job ref: 2698327df7e44d10875e3a4c5b3fa970
Location ref: Danesbury, Welwyn Hatfield
Full Job Description
Join the small-molecule team within AI for Drug Discovery (AI4DD), formerly Prescient Design, at Roche and Genentech's Computational Sciences Center of Excellence as a Machine Learning Scientist / Senior Machine Learning Scientist building agents for applied small-molecule drug design. You will develop autonomous, LLM-driven agentic workflows that orchestrate ML models, physics-based methods, and cheminformatics tools to accelerate discovery, working with world-class chemists and structural biologists.,
- Design, build, and apply agentic workflows and ML models for key challenges in small-molecule drug design.
- Fine-tune foundation models for drug discovery relevant topics using internal and external datasets and tools.
- Optimize agent-derived hypotheses in close collaboration with world-class computational and medicinal chemists and structural biologists.
- Drive scientific impact through publications, open-source releases, and conference talks.
- Collaborate widely with computational and experimental researchers at Roche and with academic partners.
You are experienced developing LLM-driven agents for scientific workflows and you understand how to orchestrate tools and models reliably. - You bring strong machine-learning foundations in linear algebra, probability and optimization, with hands-on experience with GNNs, sequence/language models and reinforcement learning.
- You are fluent in Python and modern agentic coding environments such as LangChain, ML frameworks such as PyTorch or JAX, as well as cheminformatics toolkits like RDKit or OpenEye.
- You hold a PhD or equivalent research depth in machine learning, computer science, chemical engineering or a related quantitative field such as physics or statistics.
- You have a record of scientific excellence evidenced by journal and conference publications or a public portfolio of relevant projects (e.g. hosted on GitHub/GitLab).. Preferred:
- Hands-on experience orchestrating multi-tool or multi-agent scientific pipelines.
- Hand-on experience working along the small molecule drug discovery value chain and an excitement to engage with chemists
- Familiarity with structural biology datasets