Embodied AI: building a 0→1 team

Background

The client had just closed a new round and needed a core team spanning hardware, algorithms and product within a very short window. The founder had deep technical strength but limited organisation-building experience, and competitors were chasing the same people at the same time.

Challenge

• The roles were not standard JDs — they demanded a combined judgment of technology route, product definition and team stage.
• Candidates were scattered across robotics companies, autonomous-driving teams, big-tech embodied labs and university labs.
• The client did not want the funding or product roadmap exposed early, so communication had to be paced carefully.

What we did

• A technical consultant worked through the technology route and product definition with the founder and CTO, turning “who we need” into a judgeable standard first.
• Used the AI talent graph to locate three profiles: embodied-model algorithm lead, hardware and mass-production lead, product and scenario lead.
• Ran double-blind conversations: candidates saw the business and team first, the client saw anonymised profiles first, and identities unlocked only when both sides agreed.
• The business expert committee joined the calibration for critical roles, so screening was not driven by big-company pedigree alone.

Timeline

1. Week 1Business diagnostic: breaking down technology route, product definition and team stage.
2. Weeks 2–3Talent mapping and anonymous outreach, locking onto the three core profiles.
3. Week 4First interview days, with the consultant calibrating judgment throughout.
4. Weeks 6–8Offer negotiation, expectation management and onboarding follow-up.

Outcome

• 10+Key offers
• 1 monthto close
• 3core functions

Ten-plus critical offers closed within a month, with all three core functions staffed. The talent map built during the search stayed in the library for the next hiring round.