Satlyt Launches Android for Space Computing

Rama Afullo, a Kenyan-American engineer with prior experience at Google and SpaceX, has co-founded Satlyt, a company focusing on software for orbital data centers. Satlyt announced an $8 million seed funding round, with a mission to enable multiple satellites to collaborate on large computing tasks.

Satlyt's software is designed to operate AI models directly on satellites. This approach contrasts with competitors building proprietary hardware, positioning Satlyt as a software provider offering an open ecosystem for various spacecraft. The company's technology has already demonstrated its capability to process data onboard, significantly reducing transmission costs and improving operational efficiency.

Recently, Satlyt deployed Google DeepMind's Gemma AI model on a Momentus spacecraft, achieving over a 60% reduction in error transmission size. The company's software is scheduled to launch on a SpaceX rocket on Thursday, powering a satellite from Indian startup TakeMe2Space. This mission will test cloud computing protocols for NASA and image processing for space surveillance startup Stellerian.

Looking ahead, Satlyt plans to demonstrate a shared computing system spanning two satellites next year, aiming to establish a true orbital cloud. The company anticipates that by the end of the decade, most satellites will incorporate advanced processors, and hopes to have its software operational on 20% of satellites by then. Non Sibi Ventures led the seed funding round, recognizing the potential of Satlyt's scalable business model.