Role Overview

Somewhere between the pleasing loss curve of a generative AI model for robot control and the reality of a pipette tip actually picking up liquid, there is a person with a screwdriver, a soldering iron, and a Python script held together with sheer willpower.
That person is you.
PopVax is building something unusual: a wet lab where low-cost general-purpose robotic arms, guided by learned control models, autonomously execute complex scientific experiments alongside more traditional lab automation equipment.
We have machine learning engineers building the brains. Now we need someone to build the body and more importantly, to make the brains and the body actually talk to each other.
This is the role for the engineer who lives in the seam between hardware and software. The one who can wire up a motor controller in the morning, calibrate an end-effector over lunch, write a ROS node in the afternoon, and debug a serial communication issue with an OpenTrons liquid handler before going home.

Specifically, you will:

Set up and calibrate robotic arms end-to-end. From unboxing to first successful pipetting action. Mechanical mounting, electrical wiring, sensor integration, kinematic calibration, end-effector design and attachment, workspace setup you own the entire physical stack. When the arm drifts by two millimetres and misses the well, you figure out whether its a calibration issue, a thermal expansion issue, a mounting issue, or a software issue, and you fix it.
Integrate traditional lab automation equipment. OpenTrons liquid handlers, plate readers, incubators, centrifuges, and whatever else the scientists need incorporated into automated workflows. These instruments often have their own APIs, protocols, and quirks. Youll write the software interfaces, build the physical integrations, and make them play nicely with the rest of the automation stack.
Collaborate closely with the machine learning team. The learned control models need to operate on real hardware, with real sensors, in real time. Youll be the person who ensures the hardware is set up in a way that the models can actually use the right cameras in the right positions, the right calibration, the right action spaces, the right data streaming pipeline. When the model says "move to position X" and the arm goes somewhere else, youll debug it together with the ML engineers and itll usually be a bit of both your problems.
Design and build custom hardware solutions. Off-the-shelf doesnt always cut it in a wet lab automation setting. Youll design and fabricate custom mounts, fixtures, adapters, and end-effectors. Youll work with 3D printers, basic machining tools, soldering stations, and whatever else is needed to make the physical setup work. If a bracket doesnt exist, you make it. If a connector doesnt fit, you modify it. If a sensor needs to go somewhere the manufacturer never intended, you figure it out.
Write the software that ties it all together. Drivers, control interfaces, communication protocols, state machines, monitoring scripts, safety interlocks the software layer that sits between the high-level intelligence and the physical hardware. This isnt the machine learning layer and it isnt a web app its the real-time, close-to-the-metal software that makes a robotic system actually function reliably. Python, C/C++, ROS, serial protocols, whatever the system needs.
Keep things running. Robots in labs break. Cables wear out. Calibrations drift. Sensors get splashed with buffer. Youll maintain the hardware, troubleshoot failures, and build the monitoring and diagnostic systems that let you catch problems before they ruin an experiment. Uptime matters when a robot goes down, an experiment doesnt happen.

Full-Stack Robotics Engineer

Job Description

No Referrers Available