Robotics Data Collection Operator

About Dexmate Dexmate is building the platform for physical AI — general-purpose humanoid robots and the full-stack infrastructure powering them.

Our mission is to make physical AI accessible to every developer, the way cloud computing made software infrastructure accessible before it.

We’re building systems that learn from real-world data.

The Role You will operate our robots through a wearable exoskeleton system to generate the high-quality data that trains them.

This is not a simulation role.

This is real hardware, real environments, and real task execution.

This is a contract (1099) position based on-site at our Santa Clara lab, with hours and duration aligned to active data collection campaigns.

You’ll perform structured tasks repeatedly and precisely so we can collect diverse and high-quality data for our developers.

This is a hands-on, execution-focused role where consistency and attention to detail matter.

What You’ll Do

• Teleoperate robotic arms using Dexmate’s exoskeleton system (your movements directly control the robot)
• Execute predefined manipulation tasks (picking, placing, sorting, assembly, etc.)
• Repeat tasks to generate high-volume, high-quality training data
• Follow structured task instructions and data collection workflows
• Execute variations of tasks based on engineering needs
• Provide feedback on system behavior, usability, and data quality
• Maintain safe and controlled operation in a lab environment What We’re Looking For
• Strong hand-eye coordination and spatial awareness
• High attention to detail
• Ability to follow instructions precisely and consistently
• Comfortable with repetitive, precision-focused work
• Ability to stay focused for extended periods
• Reliable and safety-conscious

Nice to Have

• Experience with VR/AR systems (Meta Quest, Apple Vision Pro, etc.)
• Familiarity with gaming or simulation environments
• Exposure to robotics, teleoperation, or similar systems

Why This Role Matters Every movement you make becomes training data.

That data is what enables robots to learn real-world tasks — from simple manipulation to complex workflows.