Simulation Engineer / Robotics

Physics simulation engineer with 5+ years of experience in contact physics, rigid-body simulation and differentiable/physics-based reward modeling for machine learning. This role owns a physics-based motion reward used to post-train generative video models via reinforcement learning: improving the dynamic and contact terms of the reward signal, optimizing it to run efficiently inside a large-scale RL training loop rather than as an offline filter, and extending it from single-body physical plausibility to human-object and human-scene interaction using 3D mesh representations.

Required skills:

• Rigid-body/contact simulation (MuJoCo, Isaac, Bullet or equivalent) *
• 3D human pose and motion representations (SMPL-family)
• Python/PyTorch with profiling and optimization of compute-heavy pipelines
• RL reward design and post-training of large generative models
• numerical methods and solver stability

Education: MS or PhD in robotics, computer graphics, mechanical engineering, computer science or a related field, or equivalent industry experience in physics simulation. -

Preferred:

• Publications or open-source contributions in physics-grounded generative modeling, human motion generation, or differentiable simulation
• Experience with generative video or diffusion models
• Experience making simulation workloads run efficiently at scale (batching, GPU acceleration, distributed training)
• Experience validating simulated physical quantities against real-world or human-judged ground truth