Research Scientist / Engineer — Geometric Deep Learning

Build models that understand irregular geometries, physical symmetries, and complex spatial relationships and make them useful for real-world engineering.

• Research and implement geometric learning architectures including graph neural networks, equivariant networks, neural operators, and neural fields.
• Develop representations for meshes, point clouds, surfaces, and volumetric simulation data.
• Build models that respect appropriate geometric and physical symmetries.
• Train on large CFD datasets and test generalization to unseen geometries and physical conditions.
• Investigate model accuracy, sample efficiency, and computational performance.
• Work with physicists, CFD specialists, and infrastructure engineers to scale and validate research.
• Contribute to technical publications and reproducible benchmarks.

• Strong foundation in machine learning, linear algebra, geometry, and numerical methods.
• Advanced Python and PyTorch or JAX.
• Hands-on experience with GNNs, geometric deep learning, neural operators, or related models.
• Experience handling non-Euclidean spatial data and independently evaluating research models.
• Degree in a relevant technical field or equivalent demonstrated expertise.

• SE(3)-equivariant architectures, Fourier neural operators, or scientific ML.
• PyTorch Geometric, e3nn, or custom GPU kernels.
• Relevant conference publications and CFD/PDE familiarity.

Validate new geometry-aware architectures that improve accuracy, efficiency, or generalization on meaningful physical benchmarks.

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