Why we are not waiting for autonomy

The standard robotics plan is to solve autonomy first and deploy second. Train in simulation, refine in a controlled space, and turn the product on once it can handle the real world unaided.

The problem is what "the real world" means for physical work. A cafe at eight in the morning is not a controlled space. Neither is a repair bench, a stockroom, or a kitchen in service. The variation is the job.

Put a person in the loop, on day one

hAire starts from the other end. A person somewhere else operates at a real address and does the work now. It is useful immediately, it is honest about what it is, and it produces exactly the data that the autonomy-first plan is waiting for: thousands of hours of a real work floor, with a competent human making the decisions.

The platform learns from that. The share of a shift that runs without intervention goes up over time, and the operator's attention moves to the parts that still need it.

What this is not

It is not a claim that autonomy is close. It is not a claim about a percentage, because we do not have a number we would be willing to defend in public yet.

It is a bet that the route to a machine doing the work runs through people doing the work, with the machine watching, rather than around them.

Status, stated plainly

Pre-launch. Deployable from mid-2026 in Amsterdam, Rotterdam and New York, in hospitality, repair, food production, maintenance, retail and small business. When that changes, this page changes, and not before.