Mecka raises $60M Series B to build the data layer for physical AI
Sequoia leads the round with NVIDIA, Qualcomm Ventures, Samsung, and Microsoft's M12 joining in. The startup's bet: robots do not lack intelligence, they lack experience — and experience is a trillion-dollar data market.
Mecka, a startup building data infrastructure for robot learning, has raised a $60 million Series B led by Sequoia Capital, the company announced on October 7. New investors in the round include NVIDIA, Qualcomm Ventures, Samsung, and M12, Microsoft's venture fund, with continued support from Kindred, Framework Ventures, and Neo. Angel investors participating include DoorDash founder and CEO Tony Xu, former ServiceNow and Snowflake CEO Frank Slootman, and former Tesla Optimus head Milan Kovac.
Mecka's pitch is that robots do not lack intelligence — they lack experience. Nobody ever wrote down how a hand grips, folds, or pours, how hard to press, or when to let go. In the company's words: motion, contact, force, and geometry are not on the internet, and cannot be scraped, licensed, or bought. Mecka describes that gap as the biggest bottleneck in robotics, calls robotics the largest productivity opportunity ahead, and predicts that data will become a trillion-dollar market.
To digitize the physical world, the company says it had to build the entire stack: capture hardware, reconstruction models, and quality infrastructure. Mecka designs and manufactures its own multi-sensor hardware, runs a global capture operation with fleets recording human demonstrations in homes and commercial environments, and maintains an internal lab that builds the computer vision and multimodal models — motion tracking, 3D reconstruction, sensor alignment — needed to turn raw, noisy reality into structured training signal. The company reports sub-centimeter hand-pose accuracy on in-the-wild data. It also offers an iOS app for capturing task-specific data and web tools for browsing, visualizing, and querying datasets.
The company tested its bet on real-world human demonstration through EgoVerse, a human-to-robot transfer study run with researchers at Georgia Tech, Stanford, UC San Diego, ETH Zurich, MIT, and Meta. The accompanying paper, first submitted on April 8 and revised on July 7, describes a collaborative platform for unifying data collection, processing, and access: 1,362 hours of human demonstrations across 80,000 episodes, spanning 1,965 tasks, 240 scenes, and 2,087 unique demonstrators, with standardized formats and manipulation-relevant annotations. The study reports that policy performance generally improves with more human data, but that effective scaling depends on alignment between human data and robot learning objectives.
On the commercial side, Mecka says it supplies several of the top frontier robotics labs and multiple Mag 7 companies, that it surpassed $100 million in run-rate revenue in June 2026, and that it projects a $300 million run rate by year-end. Those revenue figures are the company's own claims and have not been independently verified. Mecka describes itself as a modern robotics integrator: where a traditional integrator ships once and stays fixed, its deployments capture data on site, post-train on it, and improve with every hour they run — an arrangement it calls the Mecka Loop. The company does not build robots; it plugs into the hardware, model, and commercial ecosystems and owns the data, integration, and evaluation layer between them.
Mecka says the new capital scales its data infrastructure, deepens its internal research lab, and builds out commercial robot deployment. The announcement is signed by Josh, Jason, and the Mecka team, and the company says it is hiring across research, hardware, and operations. As robots move out of their cages and into warehouses, stores, and factory floors alongside people, Mecka is betting that the scarcest resource in the physical AI era is not compute or models — it is structured, real-world experience.