Research Engineer — Physics AI Evaluation & Environments

Build scientific benchmarks and simulation-backed environments that rigorously establish where physics AI models work and where they fail.

• Design and maintain evaluation suites for physics foundation models.
• Construct blind simulation test sets excluded from training.
• Compare model outputs against trusted solvers and physical references.
• Evaluate force coefficients, pressure/velocity fields, and conservation residuals.
• Measure robustness, uncertainty, calibration, and distribution shift.
• Develop environments and reward functions for agents and world models.
• Automate regression reports, reproducibility checks, and benchmarking.

• Excellent Python engineering skills.
• Numerical methods, scientific computing, and quantitative evaluation.
• Automated testing, data pipelines, and benchmark development.
• ML evaluation and reproducible experimental design.
• Ability to translate scientific requirements into dependable software.

• CFD or other numerical simulations.
• MLflow, Weights & Biases, pytest, scientific visualization.
• Uncertainty quantification, RL environments, physical conservation laws.

Make it straightforward to determine when a new model is truly better and under which physical conditions it remains reliable.

Send your CV and a relevant paper, technical project, GitHub repository, or portfolio to careers@ratiolabs.ai. Suggested email subject: “Evals and Environments”.