Physical AI Startup Humyn Labs Raises USD 47M Series A


Humyn Labs, a physical AI research and deployment company, raised USD 47 million (about 63 billion won) in a Series A round.

press-release-20261119-article-4fcbc5.jpg

Global venture capital firms Accel and Lighthouse Canton led the round. Prosus Ventures, BeeGlobal, Finserv Intelligence through Bajaj Finance, the Bajaj Alts AI Opportunities Fund, Dallas Venture Capital, Jump Trading LLC and Trifecta Capital also participated. Anita Pramoda, an angel investor and former chief financial officer of Epic, is also among the investors.

Humyn Labs records how people work in real settings such as factories, warehouses, farms and homes, using its own equipment, and processes the recordings into data for robot learning. It turns visual, audio, motion and tactile data into datasets with verified provenance and supplies them to frontier AI labs, robotics companies and world model developers.

The company also deploys robots at the same sites together with model labs and on-site operators. It measures the gap between the robots’ performance and the way people do the work, then feeds the results into the next round of model training. Its main customers are frontier AI labs, humanoid robot companies, and world model and voice AI developers. It also publishes BRIDGE, an independent physical AI benchmark series covering voice, vision and touch.

Serial entrepreneurs Manish Agarwal and Ishank Gupta co-founded Humyn Labs in 2026. Agarwal is a graduate of IIM Ahmedabad, and Gupta of The Wharton School.

Within six months of starting operations, the company’s annualized revenue topped USD 50 million (about 67 billion won). It runs data collection and processing operations directly in more than 20 countries across Latin America, East Asia, South Asia and Southeast Asia.

Humyn Labs plans to use the new funding to build robot deployment readiness infrastructure in Korea, Brazil and India, upgrade its proprietary data collection equipment and data processing engine, and expand local bases in San Francisco, London and Seoul, among other purposes.

“Physical AI can’t be better than the real-world data it learns from,” co-founder Manish Agarwal said. He said the company will use the funding to strengthen its research capabilities and build trust with AI labs and corporate customers to expand its global footprint. “A USD 10 billion (about 13 trillion won) market is opening up in data infrastructure,” he added, saying the company aims to grow into a business with USD 1 billion (about 1.3 trillion won) in annual revenue and become a global leader in the field.

Co-founder Ishank Gupta said: “Robots don’t learn work from the internet. They learn it from people who have done that work for 20 years, at the actual job site.” He said bringing model labs and on-site operators together at the same sites means every deployment feeds the training of the next model, and the company aims to make factories, farms and warehouses around the world a source of training data.

Pratik Agarwal, a partner at Accel, explained the firm’s rationale. “Physical AI is one of the technological shifts that will define the next decade, and the real competitive edge comes from data and deployment,” he said. “Humyn Labs has the technology, customer demand and drive to lead the industry.” He added that Accel led the round because it has long-term confidence in the team, and that it will work closely with the company as it expands globally.

Share

Leave a Reply

Your email address will not be published. Required fields are marked *