Get A Job.ai-World: Core Physics Engineer

What We're Building For Our Client

We are representing a confidential robotics AI organization in Paris that is developing a next-generation physics simulation platform. This open-source engine unifies multiple physics solvers—rigid bodies, finite elements, material point method, particles, cloth, and fluids—in a single environment where robotic systems can be trained at unprecedented speed and scale. Their proprietary renderer delivers real-time photorealistic output with advanced features, while their sensor suite includes cameras, lidar, IMU, contact forces, and state-of-the-art tactile simulation.

The platform runs Python-first with an in-house JIT compiler that targets CUDA, AMD ROCm, Apple Metal, Vulkan, x86, and ARM64. It supports massively batched GPU simulation for reinforcement learning and complex CPU-based scenes, running two orders of magnitude faster than real hardware while maintaining 89% correlation with physical rollouts. This is production infrastructure for physical AI that scales with compute rather than wall-clock time.

Responsibilities

You will advance the core physics capabilities of the simulation platform, delivering production-ready features that directly support internal robotics pipelines. Research is applied end-to-end: from algorithm design through implementation, testing, documentation, and deployment into live systems. Your work will be measured across five dimensions:

• Speed: Implement algorithms that enable larger stable timesteps, adaptive fidelity across spatial and temporal scales, and structure-aware solvers that allocate compute intelligently.
• Completeness: Extend the engine to cover new physics domains as use-cases demand—fluids, soft tissue, granular materials, aerodynamics, biological systems.
• Fidelity: Improve realism in contact, friction, deformation, energy dissipation, actuation, and material models.
• Versatility: Enable seamless multi-physics coupling without compromise, supporting two-way constraint-based interaction between all solvers and extensibility for custom solver integration.
• Scalability: Scale from workstation to datacenter, handling thousands of interacting entities in batched environments while preserving physical correctness.

What We're Looking For

You are both physicist and production engineer, judging methods by their real-world performance at scale. You have deep expertise in physics-based simulation with a bias toward robotics applications:

• Strong background in rigid body dynamics, finite element methods, material point method, smoothed particle hydrodynamics, incremental potential contact, position-based dynamics variants (XPBD, VBD, ABD), constrained optimization, and numerical integration of stiff systems.
• Proven track record shipping simulation code used in production—whether in a commercial engine, industrial application, or widely-adopted research codebase.
• Solid high-performance computing experience (CPU and/or GPU) with intuition for what makes numerical methods fast in practice beyond theoretical complexity.
• Rigorous validation discipline using analytical solutions, cross-engine benchmarks, and real-world data.
• Bonus: Publications at SIGGRAPH, ICRA, IROS, CoRL, or RSS; contributions to open-source physics engines.

Example challenges you will tackle:

• Implement physics level-of-detail systems that simulate high-fidelity only where agents observe and interact, coarsening elsewhere.
• Design heterogeneous environment support where each parallel world runs different models, bodies, joints, and collision geometries.
• Develop adaptive timestep control per constraint island using error-based Runge-Kutta methods and time-of-impact stepping.
• Build efficient two-way constraint-based coupling between heterogeneous solvers using ADMM and related techniques.
• Scale constraint solvers beyond current limits and unify contact resolution frameworks.
• Handle closed kinematic loops intrinsically for numerical stability without explicit constraints.

Day-to-day, you write physics code in Python that the compiler stack optimizes for every backend, and you validate against closed-form solutions, reference engines, and experimental data.

How We Work With You

Our talent team at Get A Job.ai partners with leading organizations to connect exceptional engineers with transformative opportunities. When you apply through our platform, a dedicated recruiter will conduct an initial screening to understand your background and ensure mutual fit. We then submit qualified candidates directly to the client for their interview process.

Please apply exclusively through Get A Job.ai. Do not attempt to contact the client directly, as all applications must come through our official channel to be considered.

Pay

Compensation details will be discussed during the screening process and are competitive with senior engineering roles in the Paris robotics and AI market.

Equal Opportunity: Get A Job.ai is committed to inclusive recruiting practices. We welcome applications from candidates of all backgrounds and work with client organizations that value diversity and equal opportunity.