Applied/Senior Applied AI Scientist, Translational AI Lab (TRAIL)

Roche•South San Francisco, CA

93d

About The Position

A healthier future. It’s what drives us to innovate. To continuously advance science and ensure everyone has access to the healthcare they need today and for generations to come. Creating a world where we all have more time with the people we love. That’s what makes us Roche. Advances in AI, data, and computational sciences are transforming drug discovery and development. Roche’s Research and Early Development organisations at Genentech (gRED) and Pharma (pRED) have demonstrated how these technologies accelerate R&D, leveraging data and novel computational models to drive impact. Seamless data sharing and access to models across gRED and pRED are essential to maximising these opportunities. The new Computational Sciences Center of Excellence (CoE) is a strategic, unified group whose goal is to harness the transformative power of data and Artificial Intelligence (AI) to assist our scientists in both pRED and gRED to deliver more innovative and transformative medicines for patients worldwide. The Opportunity The Translational AI Lab (TRAIL) is a team within AI Biology & Translation (AIBT) focused on applying and adapting state-of-the-art AI methods to solve key challenges in disease biology, target discovery, and translational research. TRAIL works in close collaboration with BRAID (Biology Research AI Development) and therapeutic area partners to integrate models into real-world scientific workflows. We are seeking a Scientist/Senior Scientist with a strong foundation in computational, statistical, and data science, and a passion for translating technical advances into biological and clinical impact. You’ll work across large, multimodal datasets and help evaluate, adapt, and deploy AI models to advance scientific questions in early discovery and translational contexts. Apply and fine-tune foundation models—such as large language models (LLMs), generative models, and multimodal encoders—for biological annotation, knowledge extraction, and biomarker hypothesis generation. Design workflows and pipelines that integrate model outputs with real-world biological data (e.g., gene expression, perturbation screens, clinical biomarkers). Evaluate model performance, robustness, and interpretability in collaboration with BRAID and therapeutic scientists. Build tools and interfaces (e.g., notebooks, dashboards, chat-based validation flows) that connect AI capabilities with experimental and translational use cases. Contribute to internal benchmarking, testing, and validation frameworks that enable scientific and strategic decision-making. Collaborate across diverse teams of biologists, modelers, and software engineers to translate AI capabilities into program-level insights.

Requirements

Ph.D. or M.S. (with relevant experience) in Machine Learning, Computer Science, Data Science, Computational Biology, or a related quantitative field.
0-3 years of relevant experience in AI, machine learning, or computational biology.
Strong technical foundation in deep learning and probabilistic modeling, with demonstrated project or publication experience.
Experience building and deploying ML/AI pipelines using Python, PyTorch, HuggingFace, and/or JAX; familiarity with tools like LangChain, Streamlit, or MLFlow is a plus.
Able to adapt and apply existing AI models (e.g., LLMs, encoders, transformers) in a rigorous, reproducible way to new biological domains.
Comfort working with biological or clinical data types—or strong interest in learning and collaborating closely with domain experts.
Excellent communicator who can collaborate in multi-disciplinary settings and explain technical results to scientific partners.
Ph.D. or M.S. (with relevant experience) in Machine Learning, Computer Science, Data Science, Computational Biology, or a related quantitative field.
3-6 years of relevant experience in AI, machine learning, or computational biology.
Strong technical foundation in deep learning and probabilistic modeling, with demonstrated project or publication experience.
Experience building and deploying ML/AI pipelines using Python, PyTorch, HuggingFace, and/or JAX; familiarity with tools like LangChain, Streamlit, or MLFlow is a plus.
Able to adapt and apply existing AI models (e.g., LLMs, encoders, transformers) in a rigorous, reproducible way to new biological domains.
Comfort working with biological or clinical data types—or strong interest in learning and collaborating closely with domain experts.
Excellent communicator who can collaborate in multi-disciplinary settings and explain technical results to scientific partners.

Nice To Haves

Exposure to biomedical or multiomic data (e.g., single-cell, bulk RNA-seq, CRISPR screens, protein interaction networks).
Hands-on experience with LLM-based workflows, prompt engineering, fine-tuning, or real-time retrieval and evaluation systems (e.g., RAG, AutoGen).
Experience with benchmarking, evaluation frameworks, or model interpretability in applied settings.
Prior involvement in translational research, target discovery, or biomarker identification is a plus but not required.

Responsibilities

Apply and fine-tune foundation models—such as large language models (LLMs), generative models, and multimodal encoders—for biological annotation, knowledge extraction, and biomarker hypothesis generation.
Design workflows and pipelines that integrate model outputs with real-world biological data (e.g., gene expression, perturbation screens, clinical biomarkers).
Evaluate model performance, robustness, and interpretability in collaboration with BRAID and therapeutic scientists.
Build tools and interfaces (e.g., notebooks, dashboards, chat-based validation flows) that connect AI capabilities with experimental and translational use cases.
Contribute to internal benchmarking, testing, and validation frameworks that enable scientific and strategic decision-making.
Collaborate across diverse teams of biologists, modelers, and software engineers to translate AI capabilities into program-level insights.