# Human Archive Raises Funding For Worker Datasets > Source: > Published: 2026-05-27 03:01:25.533180+00:00 # Human Archive Raises Funding For Worker Datasets Gizmodo reports that **Human Archive** has raised **$8.2 million** from investors including **Wing Venture Capital**, **NVP Capital**, **Y Combinator**, and angels from **OpenAI**, **Nvidia**, **Google**, **Mercor**, **AfterQuery**, **BAIR**, **SAIL**, **Brad Boa**, and **Meta**. Gizmodo, citing a company video and TechCrunch reporting, says the startup attaches head-mounted and wrist-mounted cameras to workers in "residential homes, restaurants, hotels, construction sites, logistics, and industrial environments worldwide." TechCrunch is reported to say Human Archive is operating **1,000** pieces of hardware. Gizmodo reports the company is headquartered in China and San Francisco. The founders are quoted describing two target datasets: a **3-D** dataset from visor-mounted cameras and a **2-D** hand-movement dataset from wrist cameras, with Raj Patel quoted as calling them "the foundational datasets required to model human sensimotor intelligence at scale." Gizmodo notes it is unclear whether footage would be used to replace workers. ### What happened Gizmodo reports that **Human Archive** has raised **$8.2 million** from investors including **Wing Venture Capital**, **NVP Capital**, **Y Combinator**, and a set of angel investors Gizmodo lists as from OpenAI, Nvidia, Google, Mercor, AfterQuery, BAIR, SAIL, Brad Boa, and Meta. Gizmodo cites a company video and TechCrunch coverage for operational details. Per Gizmodo's reporting of the company video and TechCrunch, Human Archive attaches head-mounted and wrist-mounted cameras to workers in "residential homes, restaurants, hotels, construction sites, logistics, and industrial environments worldwide." TechCrunch is reported to say Human Archive is operating **1,000** devices. Gizmodo reports the company has bases in China and San Francisco. ### Technical details Gizmodo reports the founders describe two primary dataset efforts in the company video: a **3-D** dataset captured by visor-style headsets and a **2-D** dataset focused on hand movements recorded by wrist-mounted cameras. Raj Patel is quoted in the video calling these "the foundational datasets required to model human sensimotor intelligence at scale." The Gizmodo piece frames the footage as first-person, task-level video intended for training models that capture physical interactions. ### Industry context Editorial analysis: Companies and investors have recently shown increased appetite for high-fidelity, physical-world datasets to train robotics and embodied-AI models. Observed patterns in similar data-collection efforts include rapid scale-up through partnerships with gig-economy platforms, intense scrutiny over consent and worker privacy, and regulatory attention in multiple jurisdictions. For practitioners, sourcing first-person video at scale typically raises annotation, synchronization, and calibration challenges as well as downstream dataset governance requirements. ### What to watch For practitioners and observers: monitor disclosures of partner companies or platforms that provide workers, the startup's stated data governance and consent processes, and any published dataset documentation or licensing terms. Industry observers will also watch for independent reporting on worker consent, compensation, and legal compliance where the cameras are deployed. ## Scoring Rationale The story matters because investors are funding first-person, worker-focused datasets that could accelerate embodied-AI development, while raising privacy and sourcing questions practitioners must consider. The coverage is notable but not a paradigm shift. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)