Mecka has raised a $60 million Series B led by Sequoia Capital to scale the human-motion data, sensors, reconstruction software, and deployment infrastructure used to train physical AI systems. New investors include NVIDIA, Microsoft’s M12, Qualcomm Ventures, and Samsung, while Kindred Ventures, Framework Ventures, and Neo returned.

Mecka says the round will expand its data infrastructure, internal research lab, and commercial robot deployments. TechCrunch reports the company is positioning itself as a robotics-data equivalent to the human-data businesses that helped train large language models.

Robots Have a Data Problem

Large language models could learn from enormous amounts of text already sitting on the internet. Robots do not have the same advantage. Tasks such as gripping a cup, folding fabric, opening a drawer, or repairing a machine depend on motion, contact, force, geometry, timing, and adaptation to messy real-world environments.

That is why human-motion robot training data is becoming a major infrastructure layer. BitcoinVersus.Tech has also covered sub-millimeter motion capture for humanoid training and robots learning new tasks from video. Mecka is attacking the same bottleneck by capturing real human activity at scale.

Mecka Builds the Capture Hardware Too

Mecka does not only buy videos. The company says it designs and manufactures its own multi-sensor capture hardware, then runs global collection operations in homes and commercial environments. Its research stack turns the raw data into structured signals such as motion tracking, 3D reconstruction, sensor alignment, depth, object tracking, tactile force, and point clouds.

That matters because robotics training quality depends heavily on what is measured at capture time. A normal camera clip can show what happened, but a richer synchronized sensor stack can help a model learn how it happened. That is the difference between merely recognizing a hand and reconstructing its pose, trajectory, contact, and relationship to the surrounding objects.

Egoverse Already Contains Hundreds of Thousands of Episodes

Mecka’s Egoverse project exposes the scale of the effort. Its public explorer currently lists more than 450,000 episodes, more than 4,200 hours of activity, and over 28,000 task categories spanning kitchens, workshops, labs, and homes. Tasks include sewing, preparing food, handling tools, and manipulating everyday objects.

The approach fits the broader move toward physical AI, where the model must convert perception into reliable action. That same problem appears in dexterous humanoid hands, general-purpose home robots, and industrial robots adapting to factory work.

Mecka co-founder Josh Gao also announced the round in an official LinkedIn post, saying the financing brings total funding above $120 million and will accelerate the company’s effort to build the data and deployment layer for physical AI.

The Investor List Is Part of the Story

The round pulls together companies from several layers of the robotics stack. NVIDIA supplies AI compute and robotics software. Qualcomm is pushing deeper into robotics platforms. Microsoft is investing heavily in AI infrastructure, while Samsung spans sensors, memory, semiconductors, and manufacturing.

Mecka says it already supplies several frontier robotics labs and multiple major technology companies. The company also says it surpassed a $100 million run-rate revenue in June 2026 and projects a $300 million run rate by year-end. Those are company-reported figures, but they show how quickly robot training data is turning from a research expense into a standalone business.

The Next AI Data Boom May Be Physical

The generative-AI boom created enormous businesses around text, code, images, GPUs, and data centers. Robotics may create another data market built around human demonstrations of the physical world. Unlike internet text, that data has to be deliberately recorded.

If general-purpose robots eventually work across homes, factories, warehouses, hospitals, and construction sites, the winning model may depend as much on the quality and diversity of its real-world training data as on its parameter count. Mecka’s $60 million round is another sign that investors increasingly see that data layer as infrastructure rather than a side project.

BitcoinVersus.Tech

Advertisement

BitcoinVersus.Tech covers robotics, physical AI, semiconductors, software, hardware, and the infrastructure behind modern computing.

Editor’s Note

Mecka’s revenue run-rate figures and customer descriptions are company-reported. The $60 million Series B was announced by Mecka and separately reported by TechCrunch.

We volunteer daily to help keep the information on this platform verifiably accurate. Support our independent research through the support options available on BitcoinVersus.Tech.

BitcoinVersus.tech is not a financial advisor. Content is provided for informational purposes.