Rama Afullo claims he anticipated the rise of orbital data centers well before it became a trend. The Kenyan American engineer had previously worked in Google's cloud computing division and briefly joined SpaceX's Starlink communications network in 2024. Those experiences led him to believe there was significant untapped potential for computers in space.

"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," he recounted. So he departed SpaceX and co-founded Satlyt, a company headquartered in Sunnyvale, California, and Nairobi, focused on building software for satellites. His vision is to enable multiple satellites to collaboratively handle large computing jobs in orbit.

The timing appears favorable. SpaceX has since committed fully to orbital data centers, and Satlyt's software is scheduled to launch Thursday aboard a SpaceX rocket, sharing the ride with the first prototype for Google's space data center initiative, Project Suncatcher.

Unlike SpaceX, Google, or startups such as Starcloud and Cowboy Space Company, Satlyt has no plans to manufacture its own spacecraft. Afullo's team is developing software to run AI models on satellites, an approach he likens to VMware and Snowflake — platforms that let users run complex software without concerning themselves with the underlying infrastructure.

"The folks like SpaceX who are doing orbital data centers — if they are the iPhone, we'll build Android as a horizontally integrated, open ecosystem," he explained. In essence, Satlyt aims to provide software compatible across satellites from many different companies.

Satlyt has already tested its software on two demonstration missions. The initial focus involves finding operational efficiencies for spacecraft, which typically rely on ground-based flight controllers to resolve any issues that arise. Transmitting data back to Earth — known as downlink — is costly and often slow, so using AI to make decisions onboard the spacecraft can deliver meaningful savings, whether by resolving anomalies or by processing sensor readings on the spacecraft rather than on the ground.

Earlier this year, Satlyt deployed Google DeepMind's Gemma AI model onboard a spacecraft operated by Momentus. The model reduced the size of a transmission concerning onboard software errors by more than 60% — an efficiency gain that Afullo says can save hundreds of thousands of dollars per satellite annually.

An upcoming launch will put Satlyt's software to work onboard a spacecraft built by TakeMe2Space, an Indian startup that develops computing hardware for satellites. The mission involves three customers: NASA, which is paying the company to test protocols for cloud computing in space; Stellerian, a startup focused on space surveillance (tracking objects in orbit) that wants to test image processing workloads; and TakeMe2Space itself, to demonstrate the satellite can host other companies' software.

Currently, few high-powered GPUs are in orbit. For now, the most likely opportunity lies in the kinds of workloads Satlyt is working on. But the company's next project is to demonstrate that it can create a shared computing system — a cloud — spanning two different satellites, which it expects to attempt next year. If successful, it will allow the company to build a true compute cloud in orbit as a third-party provider to spacecraft builders.

"The way to think about it is that we turn your satellite into a revenue-generating managed service," Afullo said.

Satlyt's seed round was led by Houston-based Non Sibi Ventures, where partner Bernard Harris, who served as a NASA astronaut for more than 20 years, helped ground the firm's conviction in a novel, space-based business. 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.

"We don't need data centers in space for Rama to be wildly successful, right. It can just be driven by the number of satellites going up."

Afullo's hope 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.

"If you're putting up a satellite without putting up a GPU on it, at the very least, you're doing yourself a disservice, right."

This story was originally published on October 1. It all kicks off on Tuesday. Learn more here