Embodied AI Engineer, VLA Deployment

Matter Inc • Menlo Park, CA

2d

About The Position

Matter is hiring an Embodied AI Engineer to own deployment and production hardening of VLA and learned robotic policies on its factory floor. The role bridges research-grade model performance and the reliability required in production manufacturing. The engineer will take models from the training pipeline, validate them in simulation, and deploy them on physical robots handling real parts. The goal is reliable performance under typical production conditions, including part variation, lighting changes, calibration drift, and extended shifts.

The position involves deploying VLA models on robot arms in assembly and test workcells, validating Sim2Real transfer, developing action prediction and control capabilities, and combining learned policies with traditional control methods to meet manufacturing safety and repeatability requirements. The engineer will also establish robustness protocols, instrument robots for continuous learning, and collaborate on deterministic safety constraints for probabilistic AI outputs.

Matter describes itself as building an AI-native autonomy stack for physical manufacturing. It owns and operates factories as controlled learning environments, collecting production data to train and validate AI systems on real hardware. The role offers work with live production programs, modular automation, and a Sim2Real pipeline connected to customer hardware.

Requirements

• Proven track record deploying robot learning systems on physical hardware in real environments, not just in a lab or simulation.
• Deep familiarity with VLA model architectures and the operational gap between benchmark performance and production reliability.
• Experience with ROS2-based systems, manipulation planning, and robot control stacks.
• Strong understanding of perception systems, including camera calibration, point cloud processing, and force-torque sensing.
• Ability to debug across the full stack, including model outputs, controller behavior, sensor noise, hardware faults, and communication latency.
• High tolerance for ambiguity and ability to work through problems on the factory floor.

Nice To Haves

• PhD or graduate experience in robotics, machine learning, or a related field.
• Experience with safety-critical robotic systems and functional safety standards.
• Background in manufacturing, industrial automation, or precision assembly.
• Familiarity with MARL or multi-robot coordination for factory-scale autonomy.

Responsibilities

• Deploy and productionize VLA models, such as OpenVLA, π0, Octo, or internal variants, on physical robot arms in assembly and test workcells.
• Own the Sim2Real validation workflow, verifying that models trained in NVIDIA Isaac Sim or MuJoCo transfer reliably to physical hardware before production deployment.
• Develop and tune the action prediction layer, including end-effector pose generation, trajectory planning, grasp synthesis, and force-torque feedback integration.
• Fuse learned policies with traditional control methods to meet manufacturing safety and repeatability requirements, including ISO 10218.
• Build robustness protocols for production edge cases such as part variation, lighting drift, sensor calibration, and wear-induced performance degradation.
• Instrument deployed robots for continuous learning by collecting failure cases, annotating edge cases, and feeding them back into the training pipeline.
• Collaborate with the Safety Shield layer to ensure probabilistic AI outputs are bounded by deterministic safety constraints.

What This Job Offers

• Job Type: Full-time
• Education Level: No Education Listed