Teaching a robot to fold a shirt takes hours of recorded motion, and someone has to capture, label and retarget every second of it. That bottleneck is what investors are paying for now.
Two people with knowledge of the deal told TechCrunch that Sequoia Capital is leading the financing, which is still being assembled. It would value the company at about $500M. Mecka’s product is motion: it films and analyzes how people move, then repackages that data for humanoid and other robotics programs.
The last raise came together roughly three months ago. Framework Ventures led that $60M round, joined by Menlo Ventures and SV Angel. Other robot brain developers and hardware makers have pulled in fresh money across the same stretch.
The competitive risk sits on both sides. Language models absorb text that already exists on the web, but manipulation skills must be filmed, annotated and adapted to specific hardware. Several AI labs are standing up internal motion-capture operations of their own, which would cut demand for outside suppliers over time.
Mecka has not confirmed the round, and terms can still shift before it closes.
