A startup founded by a former Google and SpaceX product manager has raised $8 million to develop software that runs artificial intelligence on satellites, according to TechCrunch. Satlyt, co-founded by Kenyan American engineer Rama Afullo, aims to create a platform that lets multiple satellites share the workload of large computing tasks in space. The company's software is scheduled to launch Thursday on a SpaceX rocket, alongside Google's first prototype for its orbital data center project called Suncatcher.

Satlyt has already flown its software on two test missions, focusing on operational efficiencies for spacecraft that typically rely on ground controllers to resolve issues. Earlier this year, the company deployed Google DeepMind's Gemma AI model aboard a spacecraft operated by Momentus, which reduced the size of a transmission about onboard software errors by more than 60%. This week's launch will place Satlyt's software on a spacecraft built by Indian startup TakeMe2Space, involving three customers: NASA, which is paying to test protocols for cloud computing in space; Stellerian, a startup focused on space surveillance that wants to test image processing workloads; and TakeMe2Space itself. The seed funding round was led by Houston-based Non Sibi Ventures, where partner Bernard Harris, a former NASA astronaut who served for more than 20 years, helped drive the firm's conviction in the space-based business.

Afullo told TechCrunch he tried to pitch the concept of orbital data centers internally at both Google and SpaceX, but both companies declined. "When I was at SpaceX, I tried to pitch this internally. They said no. When I was at Google, I tried to pitch this internally. They said no," he said. According to the report, unlike SpaceX, Google, or startups like Starcloud and Cowboy Space Company, Satlyt isn't planning to build its own spacecraft. Instead, Afullo's team is building software to operate AI models on satellites, comparing his approach to VMware and Snowflake—companies whose platforms let users run complex software without worrying about the underlying infrastructure. The report notes that using AI to make decisions on spacecraft can mean real savings, since sending data back to Earth is expensive and often slow.

The company's strategy addresses efficiencies that can save hundreds of thousands of dollars per satellite each year, whether when resolving anomalies or by processing sensor readings on the spacecraft instead of on the ground. Right now, few high-powered GPUs are in orbit, so the most likely opportunity lies in the kinds of workloads Satlyt is pursuing. The company's next project is to demonstrate it can create a shared computing system—a cloud—spanning two different satellites, which it expects to attempt next year. If successful, this will allow Satlyt to build a true compute cloud in orbit as a third-party provider to spacecraft builders. Non Sibi partner Kent Lucas told TechCrunch that one reason the firm invested was that Satlyt doesn't depend on a highest-octane vision of space data centers, which faces uncertain economics and a launch bottleneck. Afullo's goal is to be live on 20% of satellites by the end of the decade, when he expects virtually every spacecraft builder to be putting GPUs and similar advanced processors in their satellites. Satlyt is among dozens of other startups appearing at TechCrunch Disrupt as part of the Battlefield competition happening October 13–15 in San Francisco. The dual-headquarters structure in Sunnyvale, California, and Nairobi positions the company to tap talent in both markets as it scales toward becoming what Afullo calls a "horizontally integrated, open ecosystem" for satellite computing. Building software that works across many companies' satellites could reshape how the industry thinks about computation beyond Earth's atmosphere, particularly if cross-satellite cloud architectures prove viable at scale.