Humyn Labs commits $20 mn to scale human intelligence layer for Physical AI

Humyn Labs will deploy $20 million to expand human data collection and voice capabilities across global markets, supporting next-generation AI models and robotics systems

Humyn Labs, an artificial intelligence (AI) company developing human data infrastructure for physical AI firms worldwide, on Monday announced a $20 million commitment. 

 

The capital will fund data collection operations across India, South-east Asia, Latin America (LATAM), and West Asia, and is required to train the next generation of robots and voice models.

 

The firm, cofounded by Manish Agarwal and Ishank Gupta, is scaling egocentric, source-first data collection, capturing first-person and real-world human activity through visuals and movements across commercial, agricultural, and residential environments.

 

This includes how humans see, navigate, and physically interact with their surroundings, creating rich, context-aware datasets. 

 

 

Humyn Labs collaborates with frontier technology companies to convert signals from real-world communities into structured human data systems used to train and evaluate next-generation AI models. 

 

In the realm of AI, voice is emerging as a critical interface and is being used in powering real-world commands, inputs, and human-robot interactions. To support this, the company is expanding its voice data capabilities across 33 languages, dialects, accents, and code-switching patterns, ensuring that Physical AI systems can respond to human instructions with deep contextual and cultural accuracy.

 

Further strengthening this stack, Humyn Labs will also launch its Robotics Labs to build high-fidelity simulation environments and world models seamlessly integrating real-world data with training frameworks. 

 

Commenting on the developments, Agarwal said, “Every real-world AI system will need continuous human data to train and validate, making this a foundational, always-on infrastructure layer. The demand is immediate and global. As physical AI scales, we see Humyn Labs growing rapidly to serve this need at scale and becoming a key enabler of next-generation AI systems.”

 

Gupta said, “The world’s best AI models are only as good as the human intelligence behind them. Having operated across five continents, I can tell you - the Global South isn’t just where this data comes from, it’s where the future of AI is being built.” 

 

The global AI training data market is projected to reach $25 billion by 2030, with physical AI validation emerging as its most technically demanding and fastest-growing segment.