Our team is hiring a few robot engineering interns for this summer at our office in Silicon Valley.
Because these roles require significant hands-on time with physical humanoid hardware, you must be based in or willing to relocate to the Bay Area for the duration of the internship. We are open to BS, MS, or PhD candidates.
We are bypassing standard ROS 2 wrappers for our core communication layer and building a low-latency network using Zenoh, CycloneDDS, and ZMQ to drive a Unitree G1 platform. We are highly interested in students with backgrounds in RL, VLAs, or imitation learning; even though this is an infrastructure and platform role, you need to understand exactly what the data pipelines and control loops must support downstream.
What you will own:
- Direct teleoperation infrastructure development and QA testing on physical hardware.
- Platform-level engineering for the Unitree G1, focusing on maximizing pipeline determinism and minimizing end-to-end latency.
- Building and scaling data infrastructure (logging, serialization, tokenization) designed specifically to feed downstream VLA and imitation learning models.
- Optimizing and tuning Whole-Body Control (WBC) and kinematics to ensure the platform is robust enough to support future policy deployment.
What we look for:
- Strong proficiency in Python and modern C++.
- Experience or strong familiarity with RL, VLAs, foundation models, or imitation learning pipelines (giving you the foresight to build the right data infrastructure).
- Solid grounding in rigid-body kinematics, linear algebra, and control theory (WBC/IK).
- Experience with major physics engines (MuJoCo, Isaac Sim) for infrastructure validation.
If you are interested, please submit your resume and a cover letter with a brief description of the most complex robotics, simulation, or data infrastructure pipeline you have personally built. For any questions, please reach out to careers@toborlife.ai.
