At Microsoft Research AI for Science we believe deep learning has the potential to transform scientific modelling and discovery crucial for solving the most pressing problems facing society, including sustainable materials and discovery of new drugs.  

For our labs in Amsterdam (NL), London (UK), Cambridge (UK), and Berlin (DE), we seek highly motivated research software engineers to join our density functional theory (DFT) team (What is DFT?). The mission of the team is to enable highly accurate and scalable predictions of the electronic structure of molecules and materials using deep learning-powered DFT. Accurate predictions of electronic structure are essential for predictive modeling across chemistry and materials science. Learn more about our work on accurate and scalable DFT in our blog and on our project website.

The primary focus of this role will be integrating our deep learning models into high-performance DFT software frameworks, making them efficient, robust, and accessible to users. You will work on model integration, numerical algorithms, GPU acceleration, and the software infrastructure needed to bring our models into real-world scientific workflows. If you have solid software engineering expertise and enjoy bringing machine learning models into high-performance scientific software, we encourage you to apply.

This post will be open until the positions are filled.  

• Collaborate with internal and external parties on integrating our deep learning models in high-performance DFT software frameworks, targeting both CPU and GPU-based frameworks
• Prepare and maintain open-source releases and releases for beta testers
• Write custom efficient GPU implementations for our deep learning models  
• Work cross-functionally with deep learning and quantum chemistry researchers and engineers to align model development strategies with high-performance integration into CPU and GPU-based DFT software frameworks.