Silicon Valley-based startup Human Archive is leveraging India's booming gig economy to address a critical bottleneck in the robotics industry: the shortage of high-quality, real-world training data. The company partners with home services, hostel, and restaurant workers to collect egocentric video footage of daily tasks, which is then used to train robots to perform physical work. Human Archive announced on Tuesday that it has secured $8.2 million in funding from Wing Venture Capital, NVP Capital, Y Combinator, and investors including figures from OpenAI, Nvidia, and Google. Founded by four students from Berkeley and Stanford, Human Archive believes that India's vast network of gig workers offers a scalable source of the data required for advancing physical AI. The startup currently reports over 1,000 active headsets deployed across various locations. While the company declined to name specific partners initially, public discourse revealed that it faced rejection from major Indian platforms like Urban Company and Pronto. These companies cited privacy and partnership concerns, prompting a public back-and-forth between Human Archive executives and the platforms' leadership. Human Archive argues that video alone is insufficient, differentiating itself by collecting synchronized data from multiple sensors, including tactile gloves, full-body motion capture suits, and wrist cameras. This multimodal approach captures motion, force, and depth information alongside RGB-D video. To facilitate data collection, Human Archive collaborates with smaller startups that offer consumers a choice: pay a discounted rate for a service in exchange for consenting to data collection, or pay the full price for an unrecorded visit. Workers participating in the program receive a base rate of $1 per hour. Although the company maintains that its operations comply with India's Digital Personal Data Protection Act and that all data is anonymized with faces blurred, the practice has drawn scrutiny from Indian authorities regarding consent mechanisms. Despite the friction with larger industry players, Human Archive has found traction with smaller operators and is expanding its reach beyond India. The company is piloting programs in Southeast Asia and the United States, aiming to build a global platform where anyone can participate in data collection for financial compensation. Investors note that Human Archive's ability to synchronize and collect vast amounts of diverse sensor data at scale is unique. Zach DeWitt, a partner at Wing VC, stated that major labs and universities are eager to run experiments on the company's datasets. As the race to build physical AI intensifies, Human Archive's success will depend on its ability to scale partnerships and deliver the high-volume, diverse data that AI labs require to train autonomous systems effectively.