Machine Learning Scientist – Molecular Foundation Models Full-time NEW

Machine Learning Scientist

Are you a skilled machine learning scientist with experience in using the latest artificial intelligence methods to train models at scale? Do you want to make a difference in how the next generation of drug candidates are designed? Can you imagine yourself contributing to life changing science?

We are looking for an early career researcher to join the Molecular AI Department, within Discovery Sciences, where your machine learning skills will play a key role in evolving our AI capacity for molecular design. The role is located at our vibrant AstraZeneca R&D site in Gothenburg, Sweden.

If you have a few years of postdoctoral experience or have recently completed your PhD, this position will give you a superb opportunity to advance your career to the next level while influencing AstraZeneca's digital Design, Make, Test and Analyze (DMTA) platforms. Are you ready to join us?

What you will do: You will have a profound impact on early drug discovery across AstraZeneca by helping us build large, multi-modal foundation models leveraging our massive collection of internal data to help design and prioritize new molecules for synthesis.

Your contributions will include:

• Design and implement state-of-the-art methods for modeling molecular data at large scale, integrating ADMET & potency measurements, experimental structures, physics-based simulation results and more
• Develop and execute workflows for training, fine-tuning and fast inference at scale on multi-GPU systems
• Critically evaluate performance and drive data acquisition strategies using active learning to maximize accuracy
• Work closely with cross-functional project teams to apply these methods to challenging drug discovery problems
• Plan, write and publish high quality scientific papers
• Actively participate in mentoring the cohort of Postdocs, PhD students and Master Thesis students in the team

Essential requirements for the role: We believe that you have a true passion for AI, science and drug design. You have a PhD degree in a relevant computational discipline and experience in using modern machine learning methods to solve scientific problems with real impact.

More specifically you have:

• A PhD in machine learning, chemistry, physics, bioinformatics or another relevant quantitative discipline or comparable industrial experience
• In-depth scientific knowledge of modern machine learning/deep learning techniques, for example graph models, generative models, foundational models, diffusion models or transformers, demonstrated by track record of project deliveries or publications and external presentations
• Experience in Python development
• Excellent written and verbal communication skills
• Highly collaborative mindset

Desirable requirements for the role:

• Background in cheminformatics/molecular property prediction highly advantageous
• Familiar with drug discovery and/or early development in industry or in a leading academic institution
• Collaborative software development experience
• Experience with workflow automation and high-performance computing environments
• Genuine interest in driving science forward through supervising Master thesis and PhD students

Work policy: When we put people in the same room, we unleash bold thinking with the power to inspire life-changing medicines. In-person working gives us the platform we need to connect, work at pace and challenge perceptions. That's why we work, on average, a minimum of three days per week from the office. This role is located in Gothenburg, Sweden and is not available for remote work or with travel or commuting support.

Why AstraZeneca? At AstraZeneca we're dedicated to being a Great Place to Work. Where you are empowered to push the boundaries of science and unleash your entrepreneurial spirit. There's no better place to make a difference to medicine, patients and society. An inclusive culture that champions diversity and collaboration. Always committed to lifelong learning, growth and development.