We're looking for a VLA Algorithm Engineer working at the intersection of robotics, simulation, and world models to help advance our embodied AI foundation models. You'll build and optimize robotic simulation environments, develop large-scale synthetic training data pipelines, and improve how models trained in simulation transfer to real-world robots.
This role is ideal for someone who enjoys tackling challenging problems in robotics, reinforcement learning, imitation learning, and 3D vision, and wants to help build more capable, robust, and generalizable robotic systems.
Key ResponsibilitiesRobotics Simulation: Build and optimize simulation environments for robotic tasks, including manipulation, navigation, mobile manipulation, and humanoid robotics.
Embodied AI Foundation Models: Contribute to the training and optimization of embodied AI foundation models, including pretraining on large-scale egocentric human video and real-world robot data.
Multimodal Learning: Design approaches for incorporating multimodal inputs, including force and tactile sensing, and develop, train, evaluate, and optimize models that integrate these signals.
Simulation & Evaluation Infrastructure: Improve simulation and evaluation pipelines to reduce the sim-to-real gap and enhance model stability, robustness, and generalization across real-world environments.
At least 1 year of relevant professional experience or research experience in robotics, simulation, system identification, reinforcement learning (RL), imitation learning (IL), world models, 3D vision, or a related field.
Experience with robotics simulation, robotic learning, or related algorithm development.
Strong understanding of at least one relevant research area, such as reinforcement learning, imitation learning, system identification, world models, or 3D computer vision.
Simulation Platform Development: Hands-on experience developing or extending simulation platforms, with familiarity with the underlying architecture of tools such as NVIDIA Isaac Sim, Isaac Orbit, MuJoCo, PyBullet, or Genesis.
Large-Scale Synthetic Data Generation: Experience building synthetic datasets for robotics or computer vision, developing scalable data generation pipelines, or using diffusion models for large-scale synthetic data generation.
Sim-to-Real Transfer: Demonstrated experience transferring reinforcement learning or imitation learning policies from simulation to physical robotic arms or mobile robots. Familiarity with addressing simulation-to-reality differences in dynamics and observations through systematic compensation or adaptation techniques is a plus.
Health insurance
Vision care
Dental coverage
401(k)
Paid holidays
PTO (Paid Time Off)
Sick leave
Skills Required
- At least 1 year of relevant professional or research experience in robotics, simulation, system identification, reinforcement learning, imitation learning, world models, 3D vision, or a related field.
- Experience with robotics simulation, robotic learning, or related algorithm development.
- Strong understanding of at least one relevant research area, including reinforcement learning, imitation learning, system identification, world models, or 3D computer vision.
- Hands-on experience developing or extending simulation platforms such as NVIDIA Isaac Sim, Isaac Orbit, MuJoCo, PyBullet, or Genesis.
- Experience building synthetic datasets for robotics or computer vision and scalable data generation pipelines.
- Experience using diffusion models for large-scale synthetic data generation.
- Demonstrated experience transferring reinforcement learning or imitation learning policies from simulation to physical robotic arms or mobile robots.
- Familiarity with compensating for or adapting to simulation-to-reality differences in dynamics and observations.
What We Do
Maxinsights provides scalable, high-quality, human-centric training data for Physical AI and robotics foundation-model companies. Its end-to-end data solution combines a globally distributed capture network, customizable workflows, and quality-assurance processes to make training data faster and more reliable to obtain. The company’s mission is to build the global data foundation for Physical AI by delivering data at scale with unmatched quality.
