Research Engineer — Robotics Simulation & Evaluation

Description

The team is hiring a research engineer to build simulation environments, training workflows and evaluation systems for embodied AI and robotics.

The work is about creating the virtual setups and tools where robot policies get trained, tested and scored at scale before they reach physical robots. You can lean toward policy training and evaluation in simulation, or toward scene reconstruction and synthetic data, depending on your background.

This suits an engineer with solid simulation or robotics experience who likes building technically sound environments that lead to real learning gains and measurable robot improvements.

Responsibilities

• Build and maintain simulation environments used for robot learning, training and evaluation.
• Design robotic tasks, environments, reward functions and evaluation metrics.
• Use simulation platforms such as NVIDIA Isaac Sim, Isaac Lab, MuJoCo or comparable frameworks.
• Help run simulation-based policy training and automated evaluation at scale.
• Apply domain randomization and related methods to make policies more robust and easier to transfer to real robots.
• Build sensor simulation, synthetic data generation or scene reconstruction workflows, depending on your focus.
• Check simulation results against real robot behavior and look into any gaps.
• Create reusable simulation tools and pipelines that make experiments easier to reproduce and speed up research.

Requirements

• Background in robotics simulation, robotics research, simulation-based training or a closely related technical field.
• Strong Python skills and working knowledge of relevant robotics or simulation tools.
• Hands-on experience with at least one simulation platform such as Isaac Sim, Isaac Lab, MuJoCo or an equivalent.
• Understanding of robotic systems, 3D environments, physics simulation or learning-based robotics.
• Experience building working simulation environments, training workflows, evaluation pipelines or synthetic data systems.
• Strong debugging and experimentation skills, including judging whether simulation results mean anything for real-world behavior.
• Nice to have: experience training or evaluating robot policies in simulation.
• Nice to have: experience with domain randomization, sim-to-real transfer or sensor simulation.
• Nice to have: experience with scene reconstruction, synthetic data generation or physically grounded 3D environment creation.
• Nice to have: experience building reusable tools for large-scale simulation or automated evaluation.
• Nice to have: experience comparing simulation output with real robot performance.
• You may specialize in either policy training and evaluation or scene reconstruction and synthetic data. Work across both is not expected, but you must show technical depth beyond building visual scenes.

Benefits

• Health insurance
• Vision care
• Dental coverage
• 401(k)
• Paid holidays
• PTO (Paid Time Off)
• Sick leave