Mecka collects real-world human activity data, recording people performing physical tasks such as grasping, folding, and pouring using its own multi-sensor hardware, then converts those recordings into training signals for robotic systems. The company also acts as a robotics integrator, supplying the robot hardware, capturing data at customer sites, and handling ongoing operations and post-training for businesses without internal robotics teams.
Led by Sequoia Capital; new investors include Nvidia, Microsoft's venture fund M12, Qualcomm Ventures, and Samsung; Kindred, Framework Ventures, and Neo also participated; individual investors include DoorDash founder and CEO Tony Xu, former ServiceNow and Snowflake CEO Frank Slootman, and former Tesla Optimus head Milan Kovac.
Mecka's reported $100 million run-rate revenue reached just months after beginning operations, with a projected $300 million run rate by year-end, signals accelerating commercial demand for real-world robotics training data at a scale previously unseen in the sector. The round draws in major semiconductor and cloud infrastructure players, Nvidia, Qualcomm, and Microsoft, suggesting that the competition for physical AI training data pipelines is converging with the same strategic interests driving data center GPU buildout. Mecka's model of continuous on-site data capture and retraining points to an ongoing infrastructure requirement, recurring sensor fleets, compute, and storage, rather than a one-time deployment cycle.