Machine Learning Scientist

At Roche, the Machine Learning Scientist role is part of the AI Biology & Translation (AIBT) team, which focuses on advancing artificial intelligence to accelerate biomedical discovery. The company is dedicated to preventing, stopping, and curing diseases, ensuring global access to healthcare. The AIBT team combines cutting-edge AI technologies with large-scale biological data to drive innovations in target discovery, disease understanding, and translational science. The main day-to-day responsibilities of a Machine Learning Scientist at Roche involve leading the development and application of next-generation AI technologies. This includes designing DNA and RNA sequences for nucleic acid-based medicines, collaborating with stakeholders in Cell Therapy, Gene Therapy, and Vaccine Oncology, and working on diverse projects from platform development to lead optimization. The role also involves translating biological questions into predictive and generative models, and driving iterative sequence optimization in close collaboration with wet-lab experimentalists. To qualify for this role, candidates must have a Ph.D. in a quantitative discipline such as Bioinformatics, Computational Biology, Computer Science, or Machine Learning, with 0-4 years of post-doctoral or industry experience. Strong research impact at the interface of machine learning and molecular biology, evidenced by publications in top-tier journals or ML conferences, is essential. Proficiency in Python programming and deep learning frameworks like PyTorch is required, along with hands-on experience in developing, training, or fine-tuning sequence-to-function models for biological molecules, preferably DNA/RNA. Excellent communication skills in English are also necessary. Nice-to-have skills include prior experience applying machine learning in a biotech, pharma, or industry setting, and domain knowledge in nucleic acid-based therapeutics. Experience with advanced ML frameworks relevant to sequence design, such as generative modeling, active learning, and model interpretability, is also beneficial. Familiarity with high-throughput functional genomics data processing is a plus. The role offers opportunities for growth and collaboration within a dynamic team dedicated to advancing healthcare solutions. What the role asks for: - Ph.D. in Bioinformatics, Computational Biology, Computer Science, Machine Learning, or a related field - 0-4 years of post-doctoral or industry experience - Demonstrated research impact at the interface of machine learning and molecular biology - Strong Python programming skills - Proficiency in deep learning frameworks such as PyTorch - Hands-on experience developing, training, or fine-tuning sequence-to-function models for biological molecules (preferably DNA/RNA) - Excellent communication skills in English - Nice-to-have: Prior experience applying machine learning in a biotech, pharma, or industry setting