The Pulse
Mecka Raises $60 Million to Build Robot Training Data Infrastructure
Mecka announced a $60 million Series B led by Sequoia Capital to expand its robot-data infrastructure, research lab and commercial deployments. The company says it collects human demonstrations and turns them into structured data for roboti
AI.info Team · Written with AI ·
Robot hardware keeps advancing, but Mecka says machines still lack the recorded experience needed to learn physical tasks. On October 7, the company announced a $60 million Series B led by Sequoia Capital, saying it will use the money to expand its data infrastructure, research lab and commercial robot deployments.
The round brings in NVIDIA, M12—Microsoft’s venture fund—Qualcomm Ventures and Samsung, alongside existing backers Kindred, Framework Ventures and Neo. Mecka did not disclose the financing’s valuation. TechCrunch reported in September that the company was nearing a Sequoia-led deal at an estimated $500 million valuation, but said then that the terms were not final.
Mecka’s bet: record what robots cannot scrape
Mecka’s argument is that text and images do not capture the details that guide physical work: how a hand grips an object, how much force to apply, or when to release it. The company says robots need data about movement, contact, force and geometry—and that this information cannot simply be collected from the internet.
“Robots don’t lack intelligence. They lack experience,” the company’s founders and team write in the funding announcement. Mecka says it has tested the idea of using real-world human demonstrations with researchers at Georgia Tech, Stanford, UC San Diego, ETH Zürich, MIT and Meta through a human-to-robot transfer study called EgoVerse.
Hardware, capture fleets and reconstruction models
Mecka says it built a system that spans custom capture hardware, data collection and software. Its instruments use multiple sensors, selected and synchronized for the signals its models need. The company says its capture fleets record human demonstrations in homes and commercial environments.
Its internal research lab develops computer-vision and multimodal models to turn those recordings into structured data. The tasks include motion tracking, 3D reconstruction and sensor alignment. Mecka says its models achieve sub-centimeter hand-pose accuracy on data collected in uncontrolled settings.
The company says it supplies several leading robotics labs and multiple companies it describes as members of the “Mag 7.” It also claims to have passed $100 million in run-rate revenue in June 2026, a few months after starting operations, and projects a $300 million run rate by the end of 2026. Those are company figures and projections; the announcement does not provide a breakdown of customers or revenue.
From recorded demonstrations to deployed robots
Mecka is also pitching itself as a robotics integrator for businesses without their own robotics teams. It says its deployments can capture data at a customer’s site, use that data to post-train systems, and then improve as the robot operates. The new financing is intended to grow that deployment work as well as data collection and research.
The model links the company’s two businesses: collect human examples, then use the resulting data to adapt robots to specific settings. Mecka’s announcement does not name customers or identify particular deployments. Its stated plan is to expand capture instruments, research and commercial operations with the $60 million round.