Help Physical AI learn the right lesson.
Eventual builds data curation for Physical AI. We turn millions of hours of video and sensor data into the signals that decide what the next generation of models trains on.
see open positions
The problem
Physical AI models, from world models to humanoid robots, learn from millions of hours of video and sensor data. More of it, of uneven quality, arrives every day. The hard part is knowing what's in it: what happened, what's reliable and what will make the model better.
Today teams pay people to watch a sample and hope it represents the rest. Train on the wrong examples and the model doesn't just fail to improve. It learns the wrong lesson.
What we build
Data curation for Physical AI.
We process footage at petabyte scale and turn it into signals. Every signal links back to the exact moment in the source and comes with a measure of how accurate it is. There are four families:
- Actions. What happened and in what order: a failed grasp, a retry, a recovery.
- Contact and hand–object interactions. When a hand touches an object, how long it holds it, and which hand supports while the other works.
- Trajectories. How hands and objects move through the scene, frame by frame.
- Quality. Whether an example is reliable enough to learn from, or will teach the model the wrong lesson.
Labs use these signals to search, select and filter the data that goes into each training run, and as supervision targets. Signals are versioned with each data release, so each dataset is better than the last.
Who we are
From AWS, Lyft and Tesla.
We're backed by Felicis, CRV, Y Combinator and the co-founders of Databricks and Perplexity. We powered the last generation of Physical AI in self-driving. Now we're helping the next generation decide what to train on.
Our values
What drives us:
- Take pride in our work. Labs train on what we deliver. A wrong signal becomes a wrong lesson in someone's model, so we know exactly what we ship and we stand behind it.
- Create clarity. We tackle genuinely hard problems at the intersection of large-scale data systems and Physical AI. We flag ambiguity, ask tough questions, and make sure everyone understands what we're building and why.
- Be curious. Robotics, video understanding and distributed systems all move fast. We go deep on the foundations and learn from the researchers we work with.
- Own the outcome. Your work decides what the next generation of robots learns from. We move fast and measure success by how much better our customers' models get.
Benefits
We provide you with everything you need to do your best work.
- Tight-knit work environment - 4x a week in office
- Competitive compensation and startup equity
- Catered lunch and dinner for SF employees
- Commuter benefits
- Team building events & poker nights
- Health, vision, and dental coverage
- Flexible PTO
- Latest Apple equipment
- 401k plan with match