Robotic Trainer Technician (VLA / Robot Learning)
TLDR
Hands-on role building data-driven training pipelines for Vision-Language-Action models on real hardware to turn demonstrations into manipulation policies.
We are a Van Nuys robotics company training Vision-Language-Action (VLA) models to run on real hardware. We are looking for a hands-on Robotic Trainer Technician to help us collect data, teleoperate our robots, and stand up the training and evaluation loop that turns demonstrations into working manipulation policies. This is a practical, in-the-lab role: you will spend real hours driving the robot and making the data pipeline reproducible, not just reading papers.
What you'll do:
- Teleoperate and remote-operate our robots to collect demonstration data, with consistent, focused execution over several hours per day
- Train and iterate on VLA policies for manipulation tasks
- Stand up reproducible training and evaluation infrastructure (data capture, logging, eval runs)
- Run repeated trials, label outcomes, and feed results back into the next training iteration
What we're looking for:
- Hands-on experience with ML engineering or research (training and evaluating models, not just using them)
- Comfort with real-hardware robotics and repetitive, precise teleoperation
- Available onsite in Van Nuys
- Reliable, self-directed, and able to keep execution consistent over long sessions
Nice to have:
- Direct experience with VLA models, imitation learning, behavior cloning, or diffusion policies
- Familiarity with ROS, MuJoCo / Isaac Sim, or similar
- A portfolio or examples of real-robot work
Logistics:
Compensation is $20/hr. Hours are variable to start and will ramp up over time as the program scales. This is an onsite role in Van Nuys.
Sorting Robotics builds advanced robotic solutions specifically tailored for the legal cannabis industry in the US and Canada. Our technology streamlines the manufacturing processes, helping businesses increase efficiency and accuracy in a rapidly growing market.