Mecka AI Raises $60 Million From Sequoia to Gather Robot Training Data

Toronto-based robotics data startup Mecka AI has raised a $60 million Series B funding round led by Sequoia Capital, aiming to address data shortages that slow general-purpose robotics development.

Reports from TechCrunch and other industry outlets value the startup at approximately $500 million. Founded in 2024 by Josh Gao, Mogen Cheng, Jason Chong, and Duy Nguyen, Mecka previously announced a separate $60 million round led by Framework Ventures in June 2026 after preliminary talks with Sequoia surfaced in September.

Capturing Physical Movement

Training foundation models for robotics presents a distinct hurdle: physical work cannot be scraped from the web.

"Motion, contact, force and geometry aren't on the internet," Mecka noted in its announcement. "You can't scrape them, license them or buy them."

Rather than relying solely on teleoperation rigs, Mecka equips human demonstrators with wearable body sensors and smartphone cameras. These capture force, contact, and movement data across everyday physical tasks. Its EgoVerse dataset includes 1,362 hours of demonstration footage across 80,000 episodes, covering 1,965 tasks in 240 scenes recorded by 2,087 unique contributors.

Corporate Backers and Angel Investors

The round drew corporate venture arms across the tech sector, including Nvidia, Qualcomm Ventures, Samsung Electronics, and Microsoft's venture fund, M12.

Individual angel investors participating in the round include DoorDash CEO Tony Xu, former ServiceNow and Snowflake CEO Frank Slootman, and Milan Kovac, the former head of Tesla's Optimus humanoid program.

Enterprise Integration

Mecka claims its customer base includes frontier robotics research laboratories and multiple "Magnificent Seven" tech companies, though the startup has not named them publicly.

The company stated it reached a $100 million revenue run-rate in June 2026 and internally projected reaching $300 million by the end of the year, though that projection remains unverified by third parties. Alongside raw datasets, Mecka is pitching an on-site integrator model that handles sensor hardware, real-world data collection, model post-training, and continuous operations directly inside client facilities.

"That future won't be one machine. It will be an ecosystem as diverse in shape and size as the work it does," co-founders Josh Gao and Jason Chong wrote. "Soon, billions of helpful robots will work alongside us... Every one of them will have learned from a human first."