Mecka AI raises $60M to build robot-training data from human motion

Mecka AI announced a $60 million Series B round led by Sequoia, with participation from Nvidia, Microsoft’s venture fund M12, and others. TechCrunch previously reported that the startup was nearing a new funding round at a $500 million valuation.

The raise puts Mecka AI squarely in the venture spotlight as demand grows for data that can help robots learn from human behavior.

Mecka AI collects and analyzes human motion data to train humanoid robots and other kinds of robots. The company pays people to record themselves doing everyday tasks, including examples like making coffee or fixing cars, while wearing body sensors and using smartphones.

The practical takeaway: Mecka’s core product is not a robot, but the human-generated data robots need to learn physical-world tasks.

Founded in 2024, Mecka AI aims to do for robotics what Scale AI, Mercor, Surge, and other data-labeling companies have done for large language models. Those companies supply the human-generated data that AI systems learn from.

For robotics, the data challenge is physical: systems need examples of motion, manipulation, and everyday task execution, not just text or labels.

TechCrunch notes that other startups are also collecting real-world data for robot training. XDOF was in talks to raise a Series B round at a $1.2 billion valuation, according to prior TechCrunch reporting.

Human-data platforms that began with LLMs are also expanding into robotics, including Scale AI and Micro1. That signals a broader race to supply the data layer for robots that need to operate in the real world.