Mecka AI, a startup that captures and analyzes human motion data to train robots, is closing in on a $500 million valuation in a new funding round led by Sequoia Capital. The deal underscores a growing bottleneck in robotics development: the scarcity of high-quality, real-world physical data needed to teach humanoid robots and other autonomous systems how to perform everyday tasks.

The Canadian-founded startup, which launched in 2024, is moving fast. Just three months after raising $60 million from Framework Ventures in a round that included Menlo Ventures, SV Angel, and Kindred Ventures, Mecka is now preparing this significantly larger investment from one of Silicon Valley's most influential venture firms. The exact size of the new Sequoia round has not been disclosed, and terms remain subject to change.

Why Is Robot Training Data Such a Valuable Problem?

Mecka AI's founders recognized something fundamental: building general-purpose robots requires vast amounts of real-world data showing how humans actually move and interact with their environment. The startup pays people to record themselves performing everyday activities like making coffee or repairing cars using body-worn sensors and smartphones. This "egocentric" approach captures the perspective and motion patterns that robots need to learn from.

The company is essentially doing for robotics what data-labeling firms like Scale AI and Surge have done for large language models (LLMs), which are AI systems trained on massive amounts of text to understand and generate human language. By systematizing the collection and analysis of human motion data, Mecka is addressing what many robotics companies and AI labs have identified as their primary constraint: the lack of diverse, high-quality physical-world training examples.

The timing is significant. As of early June 2026, Mecka was projecting it would end the year at an annual revenue run rate of $100 million, according to co-founder Josh Gao. That trajectory suggests the market for robot training data is expanding rapidly as more companies pursue humanoid and autonomous robotics projects.

Who Are Mecka's Founders and What's Their Background?

Mecka AI was co-founded by four entrepreneurs with unconventional backgrounds for a robotics company. Josh Gao and Mogen Cheng, both Canadian, previously built a fintech startup focused on the restaurant industry. Jason Chong joined Coinbase after the cryptocurrency exchange acquired his crypto platform. Duy Nguyen handles operations at the startup.

None of the founders have robotics expertise, but they shared a crucial insight: they identified the data scarcity problem as the primary bottleneck holding back progress in general-purpose robotics, including humanoids. Rather than trying to build robots themselves, they chose to solve the upstream problem by becoming the data infrastructure layer that robotics companies depend on.

How Is Mecka AI Positioning Itself in a Crowded Market?

Mecka is not alone in recognizing the opportunity. Other startups are also collecting real-world data for robot training, including XDOF, which TechCrunch reported was nearing a new funding round at a $1.2 billion valuation. Additionally, broader human-data platforms like Scale AI and Micro1 are expanding beyond their original focus on language model training to include robotics applications.

What distinguishes Mecka is its focus on egocentric motion capture and its rapid scaling. While the startup has not publicly disclosed its customer list, many robotics companies and AI labs rely on the type of real-world data Mecka collects, alongside other methods like teleoperation, to train their models. The company's projected $100 million annual run rate by the end of 2026 suggests it has already secured significant customer traction.

Steps to Understanding Mecka's Role in the Robotics Ecosystem

• Data Collection Method: Mecka pays individuals to record themselves performing everyday tasks using body sensors and smartphones, capturing human motion from a first-person perspective that robots can learn from.
• Market Position: The startup serves as infrastructure for robotics companies and AI labs that need high-quality physical-world training data, similar to how Scale AI provides data for language model development.
• Growth Trajectory: Mecka projected $100 million in annual revenue run rate by the end of 2026, just two years after its 2024 founding, indicating rapid market adoption and customer demand.
• Competitive Landscape: Other data collection startups like XDOF and platforms like Scale AI and Micro1 are also pursuing robotics training data, creating a competitive but growing market segment.

The Sequoia investment signals confidence that the robotics industry's data hunger will only intensify. As humanoid robots and autonomous systems move from research labs into real-world applications, the demand for diverse, high-quality training data will likely become a critical competitive advantage. Mecka's ability to scale data collection while maintaining quality could position it as a foundational layer in the robotics economy.

For investors and industry observers, the $500 million valuation reflects a broader trend: venture capital is increasingly betting on the infrastructure and data layers that enable AI and robotics breakthroughs, not just the end-user applications themselves. Sequoia's participation underscores how seriously the venture world is taking the challenge of building the data pipelines that autonomous systems require to function reliably in the physical world.