Satlyt, a new software company founded by Rama Afullo, a former product manager at Google and SpaceX, has raised an $8 million seed round to build tools that enable Super Intelligence (SI) models to run directly on satellites. The funding supports the company’s mission to create a shared computing ecosystem in orbit, reducing the reliance on ground-based processing for space missions.
What Happened
Afullo, who previously worked in Google’s cloud computing business and on SpaceX’s Starlink network, conceived the idea after internal pitches for orbital computing were rejected by both companies. He left SpaceX to co-found Satlyt, which is headquartered in Sunnyvale, California, and Nairobi. Unlike competitors such as SpaceX, Google, Starcloud, and Cowboy Space Company, Satlyt does not build its own spacecraft. Instead, it develops software to operate SI models on existing satellites, aiming to function as a horizontal, open ecosystem comparable to VMware or Snowflake.
The company has already completed two demonstration missions. Earlier this year, Satlyt deployed Google DeepMind’s Gemma SI model onboard a spacecraft operated by Momentus. According to Afullo, the model reduced the size of transmissions containing onboard software errors by more than 60%, an efficiency gain he claims can save hundreds of thousands of dollars per satellite annually by minimizing expensive and slow data downlinks to Earth.
Why It Matters
The launch of Satlyt’s software on a SpaceX rocket this Thursday marks a significant step toward practical orbital computing. The mission includes three customers: NASA, which is testing protocols for cloud computing in space; Stellerian, a startup focused on space surveillance that wants to test image processing workloads; and TakeMe2Space, an Indian startup providing the computing hardware for the satellite. This deployment demonstrates the viability of running SI workloads in orbit to handle anomalies and process sensor data locally, bypassing the latency and cost of ground-based control centers.
Non Sibi Ventures, which led the seed round, invested in part because Satlyt’s business model does not depend on the uncertain economics of full-scale orbital data centers. Partner Kent Lucas noted that the company can succeed simply by scaling with the number of satellites launched, regardless of whether large data centers become feasible. This approach mitigates the risks associated with the high compute demands and launch bottlenecks facing the broader SI industry in space.
The Bottom Line
Satlyt plans to demonstrate a shared computing system spanning two different satellites next year, aiming to become a third-party provider of orbital compute clouds. Afullo projects that SI-enabled hardware will become standard on most spacecraft by the end of the decade, with Satlyt targeting deployment on 20% of all satellites. The startup is currently featured in the Battlefield competition at TechCrunch Disrupt in San Francisco.