Mecka AI raises $60M to train robots with human data from body sensors and iPhones
The AMW Read
Novelty 2: introduces a new entrant in robotics data with an unconventional human-sensor approach; Significance 2: targets a critical data bottleneck and claims $100M ARR, potentially reshaping the robotics data layer.
Mecka AI raises $60M to train robots with human data from body sensors and iPhones
Mecka AI, a New York-based startup founded by four non-traditional AI founders with backgrounds in fintech, crypto, and sneaker reselling, has raised $60 million across two undisclosed rounds ($25M Series A closed in November plus $35M follow-on) led by Framework Ventures, with participation from Menlo Ventures, SV Angel, Kindred Ventures, and angel Ted Xiao (ex-Google DeepMind, founding member at Bezos' Project Prometheus). The company collects physical motion data from humans using custom body sensors and iPhones, then uses that data to train robots, bypassing the more common teleoperation approach. Mecka AI projects a $100 million annual run rate from already-signed contracts and operates from a 40-person headcount.
Why it matters: The round represents a bet on a recurring pattern in robotics — the race for embodied AI training data as the critical bottleneck. Unlike the hyperscaler-distribution moat that defines foundation-model labs, Mecka is pursuing a data-infrastructure-first strategy, positioning itself as the "data layer" for physical AI rather than a full-stack robotics company. This mirrors the acqui-licensing pattern seen in autonomous driving (Wayve's visual training from car cameras) and recent data-scraping experiments (MicroAGI's home-cleaning recordings), but Mecka claims first-mover advantage in scaling human motion capture direct-to-robot training. The $100M ARR projection, if accurate, would make this one of the fastest revenue ramps in the robotics data segment — though the company has not disclosed customers to validate that figure.
The founding team's trajectory is notable: all four cofounders come from outside the traditional robotics or AI labs (fintech, Coinbase acquisition, sneaker flipping), a talent-dynamics signal that the robotics data layer is attracting capital and entrepreneurial attention from non-incumbents. The inclusion of Ted Xiao, a founding member of Bezos' Project Prometheus, as an angel adds a direct link to one of the largest sovereign-AI ambitions in robotics. Whether Mecka can sustain its lead against incumbents like Wayve or new entrants backed by the hyperscalers — Google, Amazon, and Tesla all have internal human-data-for-robot efforts — will determine if this capital-composition arc converges on a standalone data company or becomes an acquisition target.
