Tech Lead, Robot Control

Company Description

Roboligent is an Austin-based robotics company building Robin — a force-controlled bimanual mobile manipulator for warehouse and industrial work. Our patent-pending actuators give robots human-like sensitivity, so they work alongside people instead of behind cages. Robin is deployed with real customers today, on real production floors.

The engineering team is small and the scope is focused. Work ships to real robots in real customer environments within days, not quarters.

Role Description

You would own the robot's control stack end to end: the joint torque loop and drive firmware at the bottom, kinematics and impedance/Cartesian control above that, and the SDK that our learned manipulation policies sit on. No one currently owns this stack — the engineers who built it have left.

You would be the first hire on a control team; we expect to add one or two more controls engineers after our Series A, and this role would help shape and lead that group. This is a hands-on-hardware role — a significant part of the work is bench debugging with an oscilloscope and a real arm, not architecture work done at a whiteboard.

What you'll own

At the joint and the drive:

• The joint torque loop and its behavior at every joint on every configuration we ship
• Drive firmware — codebase, builds, versioning, flashing, and hardware-in-the-loop validation
• Calibration, commutation, and bring-up of new arm configurations
• Functional safety at the drive — Safe Torque Off, safe brake control, and the evidence behind our Declaration of Conformity

At the arm:

• Kinematics, dynamics, and impedance/Cartesian control — how the arm yields on contact and stays stable
• The control SDK — both the code and its written specification
• The control-side failure backlog, closed with root causes rather than workarounds
• Gains and control parameters as versioned, controlled artifacts — a defined tuning procedure and a validated set per arm configuration

What this role is not

• You are not a ROS 2 or motion-planning engineer. Behavior trees, navigation, teleoperation, and the autonomy stack belong to other people; you need to know where your layer meets theirs, not work inside it.
• You are not on the AI side. You will not be training models. Learned manipulation policies depend on your layer behaving predictably.
• You are not a technician. Flashing boards and running calibration are part of the job, but the goal is to turn them into procedures other people can execute.
• You are not expected to do all of this at once — see the first-six-months priorities covered in the interview process.

Qualifications

• Masters in robotics, controls, or mechatronics plus 5–8 years building products where the hardware working was your problem — a doctorate in compliant/force control with industry time, or a decade of industry with no graduate work, is equally good
• Direct ownership of both a low-level torque/current loop and a higher-level Cartesian, impedance, or admittance layer
• Some embedded or firmware work on a motor drive, servo controller, or comparable real-time target
• Substantial recent hands-on hardware time — this role gets more hands-on leverage over time, not less
• Clear technical writing — we will ask to see a specification, root-cause analysis, or test report you wrote
• Real experience developing with AI coding agents as part of your workflow
• Willing to be in the lab with the hardware — not remote-eligible; the work is physically located at the robot
• Degrees are useful signal, not a requirement

Location and logistics

On-site in Austin, Texas. Not remote, not hybrid. The hardware is here.

Occasional domestic travel to customer sites when a failure needs eyes on the machine. Not a travel-heavy role.

Process

About three weeks end to end, with feedback after every stage: an intro call with our CEO, two writing samples (a specification/root-cause/test report and an agent configuration file), a 60-minute technical conversation, a paid 3-hour working session on our bench with a real seeded fault ($750 flat, paid regardless of outcome), final conversations, and references.