Hub provides real-world training data to frontier AI labs and robotics companies. We went to 7-figure revenue agreement on data delivery during the last weeks, working with some of the biggest names in AI and robotics.

Why now

The gap between frontier models is closing, and internet-scale data is largely exhausted. The only edge left is real-world data that humans have to produce. Every lab now wins or loses on its edge cases, and edge cases need specific data captured by specific people: data that doesn't exist anywhere online. Whoever can go collect the right data first ships the better model first. Human labor is half of global GDP, and almost none of it has been recorded yet.

What Hub does

We capture this data through a global network of humans, across all continents. We route and activate the exact people who fit a lab's brief and deliver in days, not the months it takes to source a generic dataset off the shelf. Because we collect the data ourselves, we can attest its provenance, which you can't do for data you didn't capture. Our ML and QA pipelines then verify and enrich it, so we meet the hardest dataset requirements labs throw at us. The moat isn't the data, it's how fast we can go get it.

Traction

From 5-figure to 7-figure in revenue agreement on successful data delivery. We now work with some of the biggest frontier AI labs and robotics companies in the world.

The ask

Two asks: —> You're building frontier models or robots and data is your constraint.
—> You're a top robotics engineer, a PhD whose research touches robotics or egocentric data, or someone with deep experience at a leading robotics lab, we're hiring.

Reach out here: founders@hub.xyz.