Satlyt tests software to turn satellites into orbital data processing networks

Satlyt’s software enables orbital task distribution

Satlyt, a startup founded in 2024, has completed a test of software designed to transform satellites into a network of space-based data processors. According to TechCrunch, the software demonstrated its ability to distribute computing tasks among orbital computers, adjusting for communication delays and available resources. The test was part of a NASA-funded research project conducted in collaboration with the University of Houston.

The software is already deployed on three satellites in orbit. However, the test itself used only one physical computer aboard a spacecraft, with three additional simulated computers to create a network of four. The goal was to show how the software could generate and execute plans in orbit, though the team identified a trade-off in computing power requirements.

How the test worked and what it revealed

The test simulated a network of four computers, each assigned distinct roles for four users and eight services. Satlyt’s software successfully generated plans on the hardware in orbit, proving the concept’s feasibility. However, the team discovered a time-referencing issue in their software, where delays were stamped as arrival times rather than elapsed time, which could affect synchronization in a real-world deployment.

While the test marks a step toward Satlyt’s vision of virtual AI data centers in space, NASA has not yet selected the company for any specific mission or deployment of its technology.

Satlyt’s purpose and origins

Satlyt aims to build software that connects multiple satellites to share computing assignments, enabling AI and other applications to run onboard. This approach could process data closer to its source, reducing the need for ground-based infrastructure. The company participated in TechCrunch Disrupt Battle this year, finishing in the top five.

Founded by Rama Afullo, a former Google and SpaceX employee, Satlyt’s team includes engineers Leina Moli and Junn Wangari, as well as CTO Nelson Psenjen. Afullo’s background in space and software development underpins the company’s focus on solving orbital computing challenges.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Satlyt’s test is a proof point for an emerging category: software-defined satellite infrastructure. For founders and operators, this shifts the economics of space-based data processing. Instead of relying on dedicated, high-cost orbital hardware, the company’s approach repurposes existing satellites into a distributed computing network. This could lower the barrier for startups or researchers needing real-time data processing in orbit—such as climate monitoring, disaster response, or defense applications—without the capital outlay of launching bespoke hardware.

The NASA funding and partnership with the University of Houston also signal institutional validation for the concept, even if a commercial deployment isn’t imminent. For teams working on edge computing, AI in constrained environments, or satellite software, Satlyt’s work offers a benchmark: it shows what’s possible with today’s hardware but also highlights the trade-offs, like computing power overhead, that software alone may not overcome. If this model scales, it could spur demand for more capable onboard processors or new satellite designs optimized for distributed computing, creating opportunities for hardware startups in the space supply chain.

Related on Lazyfounder

• Google
• SpaceX
• More technology news