Teams at Gritt Robotics

Intern will build and operate training, data, and evaluation infrastructure: manage GPU clusters, orchestration, and large-scale data pipelines; instrument, monitor, and harden systems; test on robots; and deliver an end-to-end scoped project with mentorship. Opportunities to publish and attend conferences for qualifying interns.

Develop and evaluate robot perception components on real-world data: build datasets and evaluation harnesses, run experiments, integrate modules into the live robot stack, and demo on hardware. Work spans pose estimation, VLMs/VLAs, 3D reconstruction, object detection, and multimodal learning; opportunity to publish and attend conferences.

Intern will train and deploy robot policies (imitation learning, RL, or fine-tuned foundation models), build data collection and evaluation pipelines, run ablations, and iterate on real-hardware performance. Work includes pose estimation, vision-language and multimodal methods, 3D reconstruction, object detection, and testing policies on office robots.