Human Archive puts $8.2M behind a bigger bet on robotics training data

Human Archive, a Mountain View, Calif.-based startup focused on robotics training data, has raised $8.2 million in seed funding. The company said the round was backed by Wing Venture Capital, NVP Capital, and Y Combinator; it did not identify a lead investor.

Human Archive describes itself as "a data lab modeling human embodied intelligence." In practical terms, it collects, labels, and synchronizes video, sensor, audio, and long-horizon data for robotics and research use. The company says it already works through a network of more than 100,000 contributors and more than 500 industry partners across homes, hotels, restaurants, agriculture, industrial, construction, and retail settings.

That positioning matters because Human Archive is not presenting itself as another robotics maker. Its pitch is that better automation starts with better human data: how people think, move, and interact with the physical world. On its website, the company frames that as foundational infrastructure not only for automating manual labor, but also for research into the brain, cognition, spatial computing, and prosthetics.

The company says it was founded by Stanford and Berkeley researchers and is already offering off-the-shelf data across RGB, RGB-D, audio, tactile, and motion-capture formats. That gives a clearer picture of where it wants to sit in the market: as a data supplier for robotics and embodied AI teams that need multimodal, real-world training inputs rather than a single narrow dataset.

Human Archive did not publish a detailed use-of-proceeds statement alongside the announcement. But its website shows an active hiring push across hardware, software, and operations, including roles in engineering, firmware, camera systems, infrastructure, manufacturing, and research. That staffing mix suggests the company is building both the systems to capture data and the operational machinery to scale it.

The broader claim behind the raise is straightforward: despite decades of research, Human Archive argues that people still barely understand how humans think, move, or interact with the world. Its bet is that a larger, synchronized archive of embodied human behavior can become core infrastructure for the next wave of robotics and physical AI.