Satlyt raises $8M to run AI software across satellites, targeting 20% of spacecraft by 2030
Satlyt’s software has flown on two demo missions, including a Momentus spacecraft where Google DeepMind’s Gemma cut onboard error-data transmissions by more than 60%. A SpaceX launch will next carry it on a TakeMe2Space satellite serving NASA and others.
Satlyt, a startup building software to run artificial intelligence on satellites, announced in October 2026 that it has raised an $8 million seed round. The Sunnyvale, California and Nairobi-based company plans to deploy its platform across satellites from multiple spacecraft builders, rather than launching its own hardware.
Rama Afullo, a Kenyan American engineer who previously worked in Google’s cloud computing division and at SpaceX’s Starlink network, co-founded Satlyt after pitching the concept internally at both companies and being turned down. He describes the approach as analogous to VMware or Snowflake — software that lets operators run complex workloads without managing 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,” Afullo said.
The company has already flown its software on two demonstration missions. In one, Satlyt deployed Google DeepMind’s Gemma AI model aboard a spacecraft operated by Momentus, reducing the size of transmissions about onboard software errors by more than 60% — savings Afullo says could reach hundreds of thousands of dollars per satellite annually. A forthcoming launch on a SpaceX rocket will carry Satlyt’s software on a satellite built by Indian startup TakeMe2Space, serving customers including NASA, space surveillance startup Stellerian, and TakeMe2Space itself.
The seed round was led by Houston-based Non Sibi Ventures. Partner Bernard Harris, a former NASA astronaut, supported the investment. Fellow partner Kent Lucas noted the business does not depend on large orbital data centers to succeed: “We don’t need data centers in space for Rama to be wildly successful. It can just be driven by the number of satellites going up.”
Satlyt’s next milestone is demonstrating a shared computing system spanning two separate satellites, which it expects to attempt in 2027. Afullo’s longer-term target is to have the platform running on 20% of satellites by the end of the decade, as he anticipates most spacecraft builders will begin including GPUs and similar processors in their satellites by then.