Teams at Gradient Robotics

Train, fine-tune, and deploy vision-language-action and robot manipulation models across the full pipeline. Handle data curation, pre-training and fine-tuning, sim-to-real distillation, debugging, evaluation, and iterative autonomous improvement on physical robots.

Build and ship performance-critical, low-latency software that interfaces with robot hardware. Own end-to-end data and inference pipelines, optimize latency and determinism, and work across kernel/firmware to autonomy stack with real robots.

Work alongside founding engineers on real robotics problems: design and prototype mechanical components and test rigs, contribute to ML training and data pipelines, integrate sensors and electrical subsystems, build tooling and dashboards, run experiments, analyze results, and ship work that goes onto robots.