Spanish startup Xoople raises $130 million to map the Earth for AI

Spanish startup Xoople has closed a $130 million Series B round. Its goal is ambitious: build a constellation of satellites to systematically map the Earth for artificial intelligence systems. The company also announced a strategic partnership with American aerospace group L3Harris, which will develop and manufacture sensors for Xoople's satellites.

Demand for up-to-date geodata is coming from several fast-growing areas of AI at once. Climate monitoring, autonomous logistics, precision agriculture, urban planning and asset insurance all need not archived maps but fresh, accurate data on the Earth's surface, updated if not in real time then at least daily. Traditional geodata providers, with weekly and monthly update cycles, cannot meet that demand. This is the niche Xoople sees for itself.

The key idea is to design the product for AI from the start. The company is building small remote-sensing satellites, but it does not plan to sell images as such. The data will be structured and standardized for direct input into ML pipelines: no manual labeling, no lengthy processing. Xoople positions itself not as a stock-photo bank in orbit but as an infrastructure data provider for AI agents and analytics platforms.

The L3Harris partnership inspires confidence. L3Harris Technologies is an American defense and aerospace giant with annual revenue of about $20 billion, producing intelligence, navigation and communications equipment for the US Army, NASA and allied states. Having such a player involved in sensor production is both a guarantee of engineering standards and a potential bridge to government and defense customers, for whom the origin of components is critical.

The round size, $130 million, looks convincing even as the venture market cools. Spanish NewSpace has until recently stayed in the shadow of its British, German and French rivals. Xoople is changing that picture.

Competition in the sector is still serious: Planet Labs operates a fleet of several hundred satellites that image the entire surface of the Earth daily. But Planet Labs is first and foremost an image provider; Xoople is betting on something different: data prepared specifically for machine consumption.

The company will put the funds toward manufacturing and launching the constellation, ground infrastructure for data storage and processing, and hiring. Official announcements do not give specific dates for the first launches.

For the AI industry, the Xoople deal is another confirmation of an important trend. Compute is growing faster than the quality and freshness of data for training and inference. High-resolution, high-frequency geospatial data remains a scarce resource in many industries, from insurance and agriculture to military analysis and climate research. Whoever builds reliable infrastructure to collect it and deliver it to ML systems will be in a strong position. That is exactly what Xoople is counting on.