Claru (a Reka business line) • San Francisco • Full-time

Text has run out. The physical world - where skill lives - has never been recorded in a form a model can learn from. Claru is building that layer.

We sell data to frontier labs and robotics companies: zero to seven figures of ARR in under six months, with a team of five. It is a two-sided marketplace: labs arrive with a research problem and a budget, not a specification; contributors and partner businesses capture the work they already do, paid per approved unit. Both sides get built engagement by engagement. That is this job.

We are hiring a Program Lead and are looking for operators who are AI-native: more time with agents than with people, automation before hiring, systems that scale rather than teams that grow. That is how a team this small outruns much bigger companies. You need to be someone who can be dropped into an unfamiliar problem, with a client waiting, and come back with a plan that holds.

What you'll do

• Turn an open-ended research need into a program with a scope, a standard, a timeline and agreed economics - then own the number
• Work directly with frontier labs and robotics teams: turn a research problem into something deliverable, keep sophisticated buyers confident, and own the relationship from first call to delivery
• Partner with our supply team to build the supply side from nothing if needed: recruit, activate and retain the contributors and partner businesses a program needs, in industries and geographies where none exists
• Go at whichever blocker is in front of you: some days a spec or a tooling gap, other days a person, a partner or an incentive
• Act as product manager for the platform team: your field evidence decides what gets built next, and you speak for the contributor and the client in the roadmap
• Go where the work happens; physical operations are won by whoever showed up
• Build leverage as you go - agents, automation, playbooks - so the second program costs a fraction of the first and the tenth runs itself

A typical day

There is no standard one, but a real day might run like this: you start on the programs already live - checking with the team that supply is filming, that yesterday's batches cleared the quality bar, and chasing the sites that have gone quiet. A buyer wants to change the spec mid-program, so you work out what that does to volume, timeline and cost, and get the new version agreed in writing. Then a first call with a lab you have not worked with before: understand what they are trying to train, and be straight about what we can produce and what we cannot.

What we're looking for

• Early-stage marketplace operating experience - required. You have had to get a flywheel turning from a standing start, in a new territory or a new vertical, where neither side existed yet and you had to build both. Every engagement here is that problem again
• AI-native - required. You spend more time with agents than with people, you automate a process before you staff it, and your answer to more volume is a better system rather than more headcount
• You have operated, not only advised. Former founders, high growth early stage startup hires, consulting, or business operations background is a strong plus and exactly the rigour we want
• Structured problem solving under pressure: first principles, then a plan a client and a team can both act on
• Numerate - you build the model, find where the economics break, and argue for the change
• Credible with a research lead and with a shift supervisor alike
• Bonus: exposure to AI labs or training data pipelines