Mecka AI secures $60 million Series B for robot motion data

Mecka AI, a startup founded in 2024 that collects and analyses human-motion data for robot training, has raised a $60 million Series B led by Sequoia. Nvidia, Microsoft’s venture fund M12 and other investors also participated in the round.

The company is building a data supply operation for humanoid robots and other types of robots. It pays people to record themselves carrying out everyday tasks, including making coffee and repairing cars, while they wear body sensors and use smartphones.

Human demonstrations as robotics training material

Mecka AI’s stated ambition is to play a role in robotics comparable to that played in large language models by Scale AI, Mercor, Surge and other data-labeling businesses. Those companies provide the human-generated data from which LLM systems learn.

In Mecka AI’s model, the recordings capture human movement while a task is being completed. That makes the underlying dataset distinct from purely text-based labeling work and positions task demonstrations as a training input for robotic systems.

The financing follows reports that placed Mecka AI near a $500 million valuation. That context is reflected in Mecka AI’s reported $500 million valuation as the company moves from a reported prospective financing to an announced $60 million Series B.

A widening market for real-world robot data

Mecka AI is not alone in pursuing real-world datasets for robots. XDOF is another startup collecting training data and was reported to be in talks for a Series B at a $1.2 billion valuation.

Companies that first built human-data platforms for LLMs are also extending into robotics. Scale AI and Micro1 are among the businesses expanding their activity in that direction, placing Mecka AI in a field where data collection and analysis are becoming a defined part of the robotics supply chain.

Business implication

For organisations assessing robotic automation, Mecka AI’s funding underlines that a robot’s training data, and its fit with the tasks it must perform, should be evaluated alongside hardware, software and operational integration.