Machine Learning Engineer (MLIPs) at well-funded AI physical sciences startup

Job Title

Machine Learning Engineer (MLIPs)

Salary

Not Disclosed

Company Description

Well-funded AI physical sciences startup

Job Description

You will lead the development of machine learning interatomic potentials (MLIPs) to accelerate the discovery of advanced functional materials. By integrating ML models with first-principles physics and automated experimental data, you’ll build high-fidelity simulations that translate complex molecular dynamics into actionable insights for a physical laboratory, bridging the gap between theoretical research and industrial production.

Location

London, UK

Why this role is remarkable

• Opportunity to work at the intersection of frontier machine learning and physical sciences with a world-class team from top research institutions and industry leaders.
• Backed by elite global venture capital firms, the company is tackling high-impact climate and industrial challenges by modernizing the discovery process for critical materials.
• Unique hybrid environment where your research directly influences physical experiments in a high-throughput laboratory, ensuring your models solve real-world engineering problems.

What You Will Do

• Design and implement scalable pipelines for training and fine-tuning machine learning interatomic potentials (MLIPs) using PyTorch or JAX.
• Collaborate with physics and simulation teams to build high-quality datasets using Density Functional Theory (DFT) and optimize equivariant message-passing architectures.
• Integrate MLIP workflows into larger simulation and discovery platforms to enable rapid, large-scale screening of candidate materials for structural and functional applications.

The ideal candidate

• A PhD in Physics, Materials Science, or a related field with deep expertise in solid-state physics and computational materials modeling.
• Proven experience training MLIPs and a strong understanding of training dynamics, loss landscapes, and generalization in chemical/physical systems.
• Advanced proficiency in Python and modern ML frameworks, combined with hands‑on experience using DFT packages like VASP or Quantum Espresso.

#J-18808-Ljbffr…