Line5 builds physical AI: robots running computer vision in the field. I joined the early team as an architect and helped set the technical foundation, including the GPU vision pipeline that reads the world in front of the robot. The comms layer runs ROS 2 and Zenoh together, each doing the part it is good at.

The engineers came from backend, ML and frontend work, and each of them arrived with a settled sense of how code should be written, formed somewhere else.

So I put the shared parts where a service has to meet them in order to build at all. The common library moved into one place that every service imports, and new services started from one template instead of from whichever service happened to be nearest. Ad-hoc structures gave way to typed ones, down to physical quantity types for the navigation math, so a distance in metres could not be quietly added to one in feet. And I pointed the coding agents we had on the team at the same rules and the same lint that gate a person’s branch.

That’s what keeps a field robot debuggable when something breaks.