We’re building the foundations of decentralized, open-source AGI. As part of that mission, we’re advancing open foundation models for real-world robots—starting with the Unitree G1 as our testbed. We’re looking for interns who have already shipped code to metal, whether autonomous gait controllers, vision-language policies, or open-source robotics stacks.

You might be a fit if you have:

• Implemented control-theory algorithms, sim-to-real RL, or on-board perception that runs at the edge

• Tuned low-latency networking or distributed-training pipelines

• Thrived in messy, high-stakes problem spaces and iterated blazingly fast

What you’ll do

• Teach the G1 to run, reason, and generalize using our global compute network

• Prototype and benchmark embodied foundation models, from simulation to field tests

• Contribute code and insights to our open-source ecosystem

We’re backed by $20 million from Founders Fund, Menlo Ventures, and angels such as Andrej Karpathy and Tri Dao.

Tell us why open-source robotics matters to you, something impressive you’ve built, and how you’d accelerate decentralized AGI through robotics.