We’re building manipulation models for deploying general-purpose robots. Our data infrastructure generates rich, multi-modal human demonstration data at exponential scale. We use it to train manipulation models that work reliably across tasks, objects, and robot embodiments. A robot has no commercial value unless it works 99% of the time. Our research is focused on bridging the gap between 40% and 99% in real world deployments.\ \ **What you’ll do:** * Evaluate our models on real robots in the lab and collect failure data to improve policies * Build sim evals with varied scenes and edge cases * Collect demonstration data with our capture hardware to fine-tune policies * Set up and maintain robot arms, grippers, cameras, and sensors **You should have:** * Hands-on time with real robot hardware, even if it was just a course project or a competition * Comfort debugging across the stack, from hardware and sensors to code to the model * Experience evaluating or running policies, ideally in robotics, VLAs, world models, or imitation learning * No aversion to unglamorous work, in robotics, research and debugging have to go hand-in-hand * A track record that shows you want to see robots work in the real world, projects, research, or competitions Our culture is built on honesty and a willingness to fail fast.
If you see yourself fitting here, please apply. **Comp:** * $20-$30/hr * 10-14 weeks, beginning Jan 2027 * NYC office at Newlab, equipment budget, meals
Job details are sourced from the employer's original posting.
Open job postingAbout the company
Human Intelligence for Robots