Robotics Control Engineer

Build the control layer that turns robot intelligence into motion.

At WIM Robotics you work below the robot's application layer, on the controller that actually drives the hardware. We drive robot bodies from other manufacturers with our own controller, no original controller needed. And we open the low-level control layer that most robot makers keep closed: PID and ADRC parameters, direct drive commands over EtherCAT. If you want your algorithms to reach real motors, this is where it happens.

We just won 1st place at the AWS x NVIDIA AI Pitch Competition during a16z SF Tech Week, and WIM is an NVIDIA Inception Grand Challenge 2026 Top 10 company.

This is a full-time, on-site role in Daegu, South Korea. International applicants are welcome: we sponsor your work visa and pay for your flight to Korea and your housing deposit.

THE ENGINEERING OPPORTUNITY

Our W-RC controller runs AI and 1 kHz real-time robot control together on a single NVIDIA Jetson, over EtherCAT and ROS 2. On Jetson we hold cyclictest maximum latency under 15 microseconds, and third-party testing verified 12 axes at 1 kHz.

The work is concrete: take a policy or a motion plan all the way to the commands that move a real robot, then find out what happens on real hardware and make it better.

The same control foundation powers PLEM, our platform for commanding robots from ChatGPT and Claude, and PLEM CODE, which turns one natural-language sentence into a robot task. Your work sits exactly where those high-level instructions meet robot dynamics, timing and motion.

WHAT YOU WILL WORK ON

- Develop and validate robot control algorithms, with depth in either learning-based or classical control.

- Implement control software in C++ and Python on ROS 2 and the W-RC real-time stack.

- Take algorithms from simulation to physical robots, chase down the differences, and improve execution.

- Work across AI, motion control and embedded hardware, where timing decides whether a robot behaves.

You do not need to be an expert in both paths. Bring depth in one, reinforcement learning and imitation learning, or dynamics, state estimation, planning and force control, and learn the other side here.

THE HARDWARE IS HERE

- NVIDIA Jetson AGX Orin in our product

- NVIDIA Jetson AGX Thor development kit

- RTX Pro 6000 training server

- NVIDIA Isaac Sim and six-axis collaborative robot arms running on our own controller

WHAT YOU BRING

- Deep knowledge in at least one path: learning-based control (reinforcement learning, imitation learning) or classical control (dynamics, state estimation, planning, force control).

- C++ and Python, plus experience with ROS 2.

- International applicants: a master's degree or higher. Korean applicants: a bachelor's degree or higher.

- Willingness to work on site in Daegu.

What we read first: robots you have actually run and what broke, code or papers we can open, and a short CV with links.

WHY WIM

About 30 people from several countries, and you can work in English. Headquarters in Daegu, a Seoul office, and a US entity, WIM Robotics, Inc. NVIDIA Inception member and Plug and Play portfolio company.

WHAT WE OFFER

- Competitive salary

- Work visa sponsorship

- Company-paid flight to Korea and housing deposit

- Real robots, real customers, and a controller you can open all the way down

HOW TO APPLY

Apply with Easy Apply and your CV. If you can, add a link to code, a paper, a project or a robot demo that shows your own contribution.

The interview is one technical conversation about your own work: the problem, the decisions you made, and the evidence behind your results.

If you want to build the controller that makes an intelligent robot move, apply to WIM Robotics.