Robot Science Ops

At Generalist, we are on a mission to build general intelligence for the physical world and make it useful to everyone. This is a deeply cross-functional, execution-oriented ops role at the center of our model research and robotics stack. As a member of the Robot Science Operations staff, you will help turn robots, data, models, and evaluations into a tight, high-velocity feedback loop for model science. Foundation model intelligence is only as good as the evaluations we can hill climb on, and this role is responsible for making those evaluations real, reliable, and repeatable.

Responsibilities

Executing a range of experiments on our robot platforms; Collaborate closely with the research teams on results and be required to synthesize and interpret your findings; Designing new robotic tasks and benchmarks to evaluate model capabilities; Procuring materials and building lightweight physical benchmarks; Ensuring robots are properly configured, calibrated, and ready for rollouts and evaluations; Running structured evaluations and measuring real-world success rates; Analyzing results and closing feedback loops with ML researchers; Beta testing internal and third-party tools for teaching robots new skills; Writing clear documentation and playbooks so others can reproduce workflows; Identifying operational bottlenecks and improving system throughput end-to-end

Requirements

Be continuously diligent in the face of seemingly repetitive, but subtly changing task evaluations; Have hands-on experience with robot data collection, evaluation, or deployment; Have conceptual understanding of the full modern ML training, fine-tuning, and inference life cycles; Are comfortable running experiments and tracking real-world metrics across multiple model variants; Enjoy operating across software, hardware, and physical systems; Have some exposure to basic EE/ME tasks (wiring, mounting sensors, assembling fixtures, debugging hardware); Are highly organized and can coordinate multiple moving parts simultaneously; Write clear, structured documentation; Prefer execution and iteration speed over theoretical purity; Like being the person who “just makes it work”