Mechatronics Engineer, VLA Gripping & Assembly (Lead)

About Sorting Robotics

We create automation technology for the cannabis industry.

Tech

We operate at the intersection of software and hardware, building a full stack solution to inventory management. Our tech stack includes: Android, JavaScript, PHP, CSS, HTML, Python, AWS, mechatronics, industrial design, manufacturing, Solidworks, and embedded systems. Creating a robotic system is the easiest it has ever been, but it is still incredibly difficult because software has to blend with dynamic hardware in a seamless fashion.

The role

Sorting Robotics builds and deploys robotic systems for real manufacturing floors. We are hiring the Cognizant Engineer (CogE) to own our next flagship: a VLA-powered (Vision-Language-Action) gripping and assembly system. You will lead this product from the rig on the bench to machines running in customers facilities, and own its technical success end to end.This is a hands-on technical leadership seat. You will design the end-effector and mechatronics, integrate vision and force/tactile sensing into a 6DOF arm, and operationalize learned manipulation policies (VLA models) so the system reliably grips, places, and assembles in the real world. You work directly with the founder.**What you will own:*** The VLA gripping and assembly system cradle to grave: concept, design, build, bring-up, deployment, and field reliability.* Gripper and mechatronics design: end-effector, actuation, GD&T, rapid prototyping (3D print, laser, CNC, PCB), production-ready designs.* Perception and sensing in the manipulation loop: vision, force/torque, tactile, and the 6DOF arm.* Operationalizing VLA / learned policies: demonstration data pipeline, policy inference into real-time control, grasp/assembly evaluation, sim-to-real iteration.**You should have:*** 5+ years taking robotic systems from concept to delivery.* Identify as an AI native engineer.* Real gripper / end-effector / manipulation experience and 6DOF arm integration.* Comfort at the hardware + ML boundary: you do not have to train VLAs from scratch, but you can integrate learned policies, run inference on a robot, curate demo data, and debug why a grasp fails (ROS/ROS 2, OpenVLA / pi-0 style policies, imitation learning, sim-to-real a strong plus).* Strong Python and physical prototyping tools (CAD, 3D printing, CNC, PCB).On-site in Van Nuys / Los Angeles. Contract-to-hire or direct full-time. Comfortable with cannabis a plus (we build for the legal cannabis market).

Skills

• C++
• Python
• Machine Learning
• CAD Design