Mecka raised $60 million in Series B funding led by Sequoia Capital to scale its robotics data infrastructure, expand its research efforts, and grow commercial robot deployments. New investors in the round include NVIDIA, M12, Microsoft’s venture fund, Qualcomm Ventures, and Samsung, with continued participation from Kindred, Framework Ventures, and Neo. The round also included angel investors such as Tony Xu, founder and CEO of DoorDash; Frank Slootman, former CEO of ServiceNow and Snowflake; and Milan Kovac, former head of Optimus at Tesla.
Mecka is building what it describes as a data and deployment layer for robotics, focused on capturing the real-world human behavior and physical interactions needed to train increasingly capable robots.
The company’s thesis is that robotics development is constrained less by hardware than by the lack of high-quality physical-world data.
While large language models have been trained on enormous volumes of written and digital information, people’s everyday physical actions have never been systematically recorded in a format robots can learn from.
Those actions include how humans grip, fold, pour, apply pressure, release objects, and respond to differences in motion, force, contact, and geometry.
Mecka has built an integrated technology stack to collect and process this information.
The company designs and manufactures its own multi-sensor capture hardware, selecting and synchronizing sensors based on the requirements of the robotics models that will ultimately use the data.
Mecka also operates global capture fleets that record human demonstrations in residential and commercial environments.
Its internal research laboratory develops computer vision and multimodal AI models that transform raw sensor information into structured datasets for robotics training.
These capabilities include motion tracking, 3D reconstruction, sensor alignment, and hand-pose estimation.
Mecka said its technology has achieved sub-centimeter hand-pose accuracy on real-world data.
The company has also collaborated with researchers from Georgia Tech, Stanford University, UC San Diego, ETH Zürich, MIT, and Meta through EgoVerse, a research effort focused on transferring human demonstrations to robots.
Mecka currently supplies data and technology to several frontier robotics laboratories and multiple companies within the group commonly referred to as the Magnificent Seven.
The company said it surpassed $100 million in run-rate revenue in June 2026, only a few months after beginning operations.
Mecka expects its run-rate revenue to reach approximately $300 million by the end of 2026.
Beyond providing robotics training data, the company is expanding into commercial robot deployment.
Mecka describes itself as a modern robotics integrator that combines hardware, on-site data collection, model post-training, systems integration, and ongoing operations.
Its approach differs from traditional robotics integration by continuously collecting data after deployment and using that information to improve robot performance over time.
The company is targeting enterprises that want to adopt physical AI but do not have internal robotics teams capable of building and operating the required systems.
Mecka plans to use the Series B financing to manufacture additional data capture instruments, deepen its robotics and AI research, expand its data infrastructure, and deploy more robots in commercial environments.
The company is also hiring across research, hardware, and operations as it scales.
KEY QUOTES:
“Fifty years ago, a robot needed five hours to cross a room. Last year, humanoids raced the hundred-meter dash in a Beijing stadium. The hardware kept improving. What it was waiting on was a way to learn.”
“Language models learned to reason by reading what humanity wrote down. No human ever started from zero, either — we inherited the experience of every generation before us. Robots will inherit ours, but only if someone records it.”
“No one ever wrote down how a hand grips, folds or pours, how hard to press, or when to let go. Motion, contact, force and geometry aren’t on the internet. You can’t scrape them, license them or buy them. Robots don’t lack intelligence. They lack experience.”
Mecka team statement
