Machine Learning Software Engineer, Research

📋 Description

• • Join PhysicsX as a Machine Learning Software Engineer, Research and sit at the intersection of cutting-edge physics simulation and state-of-the-art ML. You will design, train, and deploy neural networks that accelerate multi-physics solvers by orders of magnitude, enabling real-time design and optimization of next-generation aerospace, energy, and semiconductor products.
• • Own the full ML research lifecycle—from literature review and data curation to model architecture, distributed training, and production-grade inference. You will work with terabyte-scale CFD, FEA, and quantum-chemistry datasets generated by our internal physics engines and partner labs.
• • Collaborate daily with physicists, HPC engineers, and product teams to translate physical laws into differentiable code. Expect to co-author papers, file patents, and present findings at NeurIPS, ICML, APS, and SciPy.
• • Build and maintain a scalable experimentation platform that supports hyper-parameter sweeps, meta-learning, and reinforcement learning across thousands of GPUs on AWS and on-prem clusters. You will champion MLOps best practices: containerized training, experiment tracking, model versioning, and automated CI/CD for GPU-optimized inference services.
• • Pioneer physics-informed neural networks (PINNs), neural operators, and transformer-based surrogate models that respect conservation laws, boundary conditions, and symmetries. Your work will reduce simulation times from hours to milliseconds, unlocking interactive design loops previously impossible.
• • Develop novel data-generation strategies using active learning and adaptive mesh refinement to create high-fidelity training sets with minimal compute cost. You will also curate open-source benchmark datasets that set new industry standards.
• • Create interpretability and uncertainty-quantification tools so domain experts can trust ML predictions in safety-critical applications such as rocket engine combustion, fusion reactor plasma stability, and lithography mask correction.
• • Mentor junior researchers and interns, run internal reading groups, and contribute to our technical blog. Your leadership will shape the culture of scientific rigor and reproducibility across the company.
• • Partner with external universities and national labs on joint grants and publications. You will help secure multi-million-dollar research contracts by demonstrating clear technical roadmaps and performance milestones.
• • Impact: Your models will directly influence the design of hardware that reaches Mars, accelerates the transition to net-zero energy, and shrinks semiconductor nodes below 2 nm. Every line of code you write shortens the path from scientific insight to real-world engineering breakthrough.