Data Science & AI Innovation Postdoctoral Fellow in Machine Learning for Chemical Synthesis and Reactivity Prediction

Research Opportunity

Join an interdisciplinary research project at the intersection of artificial intelligence, machine learning, and synthetic chemistry. This fellowship aims to develop next-generation machine learning approaches that predict chemical reaction outcomes, reaction conditions, and molecular reactivity using large-scale proprietary and public reaction datasets. By leveraging more than two decades of reaction knowledge generated within Novartis, the fellow will investigate how modern AI methods, including graph neural networks, transformer architectures, and foundation models, can improve the efficiency and success rate of chemical synthesis.

The project will focus on building predictive models that help chemists design more efficient synthetic routes, identify optimal reaction conditions, and expand access to diverse chemical space. The fellow will collaborate closely with experts in data science, computational chemistry, medicinal chemistry, and synthesis technology, with opportunities to integrate predictive chemistry models into generative AI workflows and emerging laboratory automation platforms. The research is expected to result in high-impact publications and contribute to accelerating the Design-Make-Test-Analyze cycle in active drug discovery projects.

Key Responsibilities

• Analyze large-scale chemical reaction datasets from proprietary and public sources to identify trends, opportunities, and challenges in chemical synthesis.
• Develop, implement, and evaluate machine learning models for predicting reaction success, reaction conditions, yield, regioselectivity, and molecular reactivity.
• Benchmark state-of-the-art AI approaches, including graph neural networks, transformer models, and foundation models, against relevant synthesis prediction tasks.
• Investigate novel pre-training strategies leveraging large-scale chemistry and physics-based datasets to improve predictive performance and generalization.
• Collaborate closely with medicinal chemists, synthetic chemists, and automation experts to address real-world drug discovery challenges.
• Apply predictive models to enable broader substrate scope exploration, reaction optimization, and library synthesis design.
• Explore integration of synthesis prediction models with generative chemistry and AI-driven molecular design workflows.
• Present research findings internally and externally, publish in leading scientific journals, and contribute to the broader scientific community.

Essential Requirements

• PhD in Data Science, Computer Science, Machine Learning, Cheminformatics, Computational Chemistry, Chemistry, Pharmaceutical Sciences, or a related quantitative discipline completed prior to the fellowship start date. The program is intended for scientists immediately following their PhD training (graduated in 2026).
• Demonstrated record of scientific achievement (publications, presentations, patents, or equivalent)
• Demonstrated experience developing and applying deep learning methods to scientific or chemical datasets.
• Strong programming skills in Python and familiarity with modern machine learning frameworks and architectures, such as graph neural networks, transformers, large language models or foundation models for scientific applications.
• Experience with data analysis, statistical modeling, and handling large, complex datasets.
• Strong commitment to learning, innovation, and professional development
• Ability to work effectively in highly collaborative, multidisciplinary research environments.
• Excellent communication skills and ability to present complex scientific concepts to diverse audiences.

Desirable Requirements

• Experience curating, processing, and analyzing large-scale chemical reaction datasets, including reaction encoding, atom mapping, reaction classification, and extraction of chemical knowledge from structured or unstructured data sources.
• Experience designing and querying relational databases or other structured data systems for scientific data management and large-scale analytics.

Why Join the Program?

The Novartis Biomedical Research Postdoctoral Fellowship Program is designed to develop the next generation of scientific leaders, powering the future of medicine, through rigorous research, and immersive learning experiences, such as implementation of AI tools in biomedical research.

Postdoctoral Research Fellows benefit from:

• Guidance from accomplished scientific leaders and subject matter experts
• Access to advanced technologies, platforms, and research capabilities
• Collaboration across disciplines and organizational boundaries
• A global and diverse community of postdoctoral fellows
• Dedicated programming designed to help fellows thrive throughout their careers.
• Personalized experiential learning opportunities through a Postdoc Practicum that empower fellows to explore new scientific domains, build cross-functional expertise, and expand their impact beyond their primary research project.
• Opportunities to present research, publish in leading journals, and build an international scientific network

This is a full-time training position of up to three years in duration.