Research Scientist, Vision-Language-Action Models

Matter Inc • Menlo Park, CA

13d

About The Position

Matter is building an AI-native autonomy stack for physical manufacturing in the United States. It operates as a contract manufacturer, deploying software and autonomy in its own factories, with the long-term vision of becoming infrastructure for American manufacturing.

The Research Scientist will lead the development and deployment of Vision-Language-Action (VLA) models for robotic manipulation in live manufacturing work cells. This is a hands-on production role involving model training, closing the Sim2Real loop, and deploying models on physical robots running production programs. Matter’s Sim2Real pipeline uses NVIDIA Isaac Sim, physics-accurate virtual builds of modular assembly equipment, and data collected from real factory operations. The scientist will help improve models through production runs.

The role includes developing and fine-tuning models for precision assembly, designing and managing Sim2Real training, evaluating real-world manipulation against manufacturing tolerances and repeatability requirements, and collaborating with controls and automation engineers to combine learned policies with traditional control architectures for production safety. The scientist will contribute to Physical AI architecture decisions covering model selection, data strategy, training infrastructure, and deployment protocols. Research publication is encouraged, but shipping production systems is the primary measure of success.

Matter offers the opportunity to validate VLA research on a production factory floor, handling real parts for customers, with feedback grounded in factory operations.

Requirements

• PhD, graduate-level, or equivalent research depth in robotics, machine learning, or a related field.
• Hands-on experience training and deploying VLA, VLM, or generalist robot policies on physical hardware, not just in simulation.
• Strong foundation in imitation learning, reinforcement learning, and general machine learning methods.
• Proficiency in PyTorch.
• Experience with NVIDIA Isaac Sim, MuJoCo, or similar physics engines.
• Ability to debug the full stack, including model architecture, training data quality, simulation calibration, sensor noise, and hardware edge cases.
• Comfort working in a high-velocity, ambiguous environment and owning systems end-to-end.

Nice To Haves

• Experience with MARL or multi-robot coordination.
• Background in manufacturing, industrial automation, or robotic assembly.

Responsibilities

• Develop and fine-tune VLA models for precision assembly tasks, including dexterous manipulation, part handling, and test operations.
• Design and manage the Sim2Real training pipeline, including domain randomization, synthetic data generation, physics simulation, and sim-to-physical transfer.
• Build evaluation frameworks to benchmark real-world manipulation performance against manufacturing tolerances and repeatability requirements.
• Collaborate with controls and automation engineers to fuse learned policies with traditional control architectures for production safety.
• Contribute to Physical AI architecture decisions, including model selection, data strategy, training infrastructure, and deployment protocols.
• Publish novel research in top-tier conferences, while prioritizing the delivery of production systems.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: Ph.D. or professional degree