Mecka AI raises $60M Series B led by Sequoia Capital to sell human motion data for robot training

The October 8 round brought in chipmaker and Big Tech venture arms alongside Sequoia: Nvidia, Microsoft's M12, Qualcomm and Samsung. The Latent reported that Mecka's annualized revenue run-rate has topped $100 million. The company pays people to record everyday tasks with body-worn sensors and phones. It then turns the recordings into force, contact and motion data that robotics teams use to train models. Robot training data is getting crowded. Munich-based MicroAGI says it hit $100M ARR six months after launching a similar paid recording platform. Simulation companies such as Axis Robotics argue synthetic data can cut the need for real demonstrations. Union Square Ventures also named physical-world data a priority in its new fund thesis this week. For founders in physical AI, the round shows that data supply, not just model design, is drawing top-tier money and chipmakers as strategic investors. Teams training manipulation policies should expect more vendors and lower prices for human demonstration data.