Open Orion

Job Summary:
Open Orion, Inc. is an early stage research and design company focused on enhancing engineering through AI and simulations. The role involves building advanced AI systems that integrate machine learning with engineering design, creating intelligent solutions for geometry and physics-related challenges.
Responsibilities:
• Design and build agentic AI pipelines for simulation and geometry workflows, systems that can autonomously set up, run, and interpret engineering simulations
• Develop surrogate models that approximate expensive physics solvers with learned representations
• Build geometry reasoning systems AI that understands CAD topology, can classify features, and make design decisions
• Create geometry generation models that produce valid, manufacturable CAD geometry from specifications, constraints, or natural language
• Integrate ML models into production simulation pipelines for design space exploration and optimization
• Enable gradient-based design optimization through differentiable geometry and physics
• Develop and maintain open-source tools that push the state of the art in AI-assisted engineering
Qualifications:
Required:
• Strong ML/AI fundamentals i.e deep learning, transformers, generative models, reinforcement learning
• Experience with CAD/CAE workflows (i.e you've worked with simulation tools, mesh generators, or geometry kernels and understand the engineering context)
• Ability to build agentic systems multi-step AI pipelines that plan, execute, and adapt
• Familiarity with computational geometry concepts meshes, BREP/NURBS representations, surface/volume discretization
• Proficiency in Python and at least one ML framework (PyTorch, JAX, TensorFlow)
• Experience building surrogate models or physics-informed neural networks
Preferred:
• Hands-on experience with CFD or FEM solvers
• Experience with geometry kernels or parametric CAD APIs
• Background in differentiable simulation or differentiable rendering
• Familiarity with mesh generation algorithms
• Experience with 3D deep learning point clouds, implicit representations, graph neural networks on meshes
• Contributions to open-source scientific computing or CAD/CAE projects
• Aerospace, mechanical, or civil engineering domain knowledge
Company:
Founded in , the company is headquartered in , , with a team of 2-10 employees. The company is currently Early Stage.