Human Archive
Human Archive collects embodied intelligence data to train robotics and physical AI.
Founded
2026Human Archive collects embodied intelligence data to train robotics and physical AI.
Human Archive differentiates through its focus on embodied intelligence data captured from real human workers in diverse physical environments rather than synthetic or lab-generated datasets.
The company leverages India's gig economy to scale data collection cost-effectively, with 1,000+ active headsets deployed and expansion into Southeast Asia and the United States underway.
Source: humanarchive.ai
The robotics and physical AI market is experiencing rapid growth as labs race to build machines capable of performing real-world tasks, creating massive demand for high-quality training data.
Human Archive is positioned to become foundational infrastructure for this market by supplying the multimodal embodied datasets that frontier AI labs need to train next-generation robotics systems.
Source: humanarchive.ai
Human Archive was founded in early 2026 by Stanford and UC Berkeley researchers Raj Patel (CEO), Samay Maini, Rushil Agarwal, and Shloke Patel.
The company raised an $8.2 million seed round in May 2026 led by Wing Venture Capital and NVP Capital, with participation from Y Combinator and angel investors from OpenAI, Nvidia, Google, and Meta.
Source: humanarchive.ai
Human Archive operates a data lab that models human embodied intelligence by collecting, labeling, and synchronizing multimodal data from real-world environments.
The platform deploys hardware including headset-mounted cameras, tactile gloves, and motion-capture suits across homes, hotels, restaurants, agriculture, and industrial facilities through a contributor network of 100,000+ participants and 500+ industry partners.
Source: humanarchive.ai