In the burgeoning landscape of the New Space economy, where the race to launch constellations of satellites has transformed the night sky, a new frontier is quietly emerging: the orbital data center. While industry titans like SpaceX and Google are pouring billions into the physical infrastructure of space-based computing, a lean, dual-headquartered startup named Satlyt is betting that the real value lies not in the hardware, but in the software layer that orchestrates it.
Founded by Kenyan American engineer Rama Afullo, Satlyt recently emerged from stealth with an $8 million seed round. The company’s mission is ambitious: to build a universal, hardware-agnostic platform that allows satellites to share computing power, essentially creating a decentralized "cloud" in orbit. As Afullo puts it, if companies like SpaceX are the Apple of the space industry—building a tightly controlled, vertically integrated ecosystem—Satlyt intends to be the Android: an open, flexible platform accessible to any satellite manufacturer.
The Genesis of an Orbital Vision
The path to Satlyt began in the boardrooms of the tech giants. Afullo, a veteran of Google’s cloud computing division and a former engineer at SpaceX’s Starlink, spent years witnessing the industry’s singular focus on connectivity. During his time at both companies, he pitched the idea of utilizing the vast, untapped computational capacity of satellites.
"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. They said no," Afullo told TechCrunch.
The rejection, however, became the catalyst for innovation. Afullo realized that while the industry was obsessed with launching more satellites, they were failing to leverage the existing ones as anything more than simple relay stations. He concluded that the industry needed a layer of abstraction—similar to how VMware or Snowflake revolutionized terrestrial data centers by decoupling software from physical infrastructure. In 2024, he co-founded Satlyt, establishing headquarters in both Sunnyvale, California, and Nairobi, Kenya, to bridge the gap between Silicon Valley’s software prowess and the global demand for space-based processing.
Chronology of a High-Stakes Pivot
The trajectory of Satlyt has been marked by rapid iteration. Unlike traditional aerospace firms that spend decades on R&D, Satlyt adopted the "move fast and break things" philosophy of the software industry.
- 2024 (Early): Afullo departs the major aerospace players to incorporate Satlyt.
- Mid-2024: The company conducts two successful demonstration missions, proving that AI models could run reliably in the harsh, high-radiation environment of Low Earth Orbit (LEO).
- Late 2024: A significant breakthrough occurs when Satlyt deploys Google DeepMind’s Gemma AI model on a spacecraft operated by Momentus. The mission demonstrates a 60% reduction in transmission sizes for onboard software error reporting, proving the tangible cost-saving potential of edge computing in space.
- This Week: Satlyt’s software takes flight on a SpaceX rocket, housed within a spacecraft built by TakeMe2Space, an Indian hardware startup. This mission serves as a critical stress test, hosting workloads for NASA, surveillance firm Stellerian, and TakeMe2Space itself.
Supporting Data: The Economics of Edge Computing
The logic behind Satlyt’s business model is rooted in a fundamental constraint of space flight: the "downlink bottleneck." Sending massive amounts of raw data back to Earth is not only slow and energy-intensive, but it is also prohibitively expensive. Currently, most satellites are "dumb" sensors, sending all their raw data to ground stations for processing.
By enabling AI processing directly on the spacecraft, Satlyt effectively turns the satellite into a smart edge device. The deployment of the Gemma model, which successfully compressed error-reporting data by 60%, is a proof-of-concept for a broader economic shift. Afullo estimates that such efficiencies can save operators hundreds of thousands of dollars per satellite annually.
The company’s $8 million seed round, led by Houston-based Non Sibi Ventures, reflects a growing investor confidence in this software-first approach. Bernard Harris, a partner at Non Sibi and a former NASA astronaut, brought the perspective of someone who has seen the limitations of orbital hardware firsthand. The investment thesis is simple: Satlyt does not need to own the expensive, high-risk hardware to be profitable. It simply needs to be the software layer that makes the hardware more efficient.
"We don’t need data centers in space for Rama to be wildly successful," says Kent Lucas, a partner at Non Sibi Ventures. "It can just be driven by the number of satellites going up."
The Implications: A Shared Cloud in Orbit
The implications of a successful "orbital cloud" are profound. Currently, high-powered GPUs are a rarity in orbit, but that is changing rapidly. As satellite builders incorporate more advanced processors to handle increasing data loads, they will face the same challenges that led to the creation of cloud computing on Earth: the need for load balancing, resource sharing, and scalable software environments.
Satlyt’s next major milestone, scheduled for next year, involves a "shared computing system" that spans two different satellites. If successful, this will mark the birth of a genuine cloud network in space, where computing tasks can be offloaded from one satellite to another in real-time, regardless of the manufacturer.
For the satellite operator, this transforms a depreciating asset—a satellite that simply carries a payload—into a revenue-generating managed service. "The way to think about it is that we turn your satellite into a revenue-generating managed service," Afullo says.
The Road Ahead: 20% by 2030
Afullo is not content with niche applications. His goal is to have Satlyt’s software running on 20% of all orbiting satellites by the end of the decade. He views the current state of satellite design as archaic, noting that launching a satellite without a GPU is increasingly becoming a strategic oversight.
"If you’re putting up a satellite without putting up a GPU on it, at the very least, you’re doing yourself a disservice," he asserts.
The tech industry’s shift toward distributed computing, AI-at-the-edge, and decentralized networks is finally reaching the stars. As SpaceX and other heavyweights focus on the physical logistics of orbital data centers, Satlyt is banking on the fact that the true revolution will happen in the lines of code that govern them.
The industry is watching closely this week. As the SpaceX rocket carries Satlyt’s software into the vacuum of space, the mission represents more than just another launch; it is a test of whether the principles of Silicon Valley—openness, modularity, and rapid software iteration—can thrive in the most hostile computing environment in existence. If Satlyt succeeds, the "Android of Orbit" may become the backbone of the next generation of space exploration, proving that in the future, the most powerful tool in space will not be the rocket, but the operating system running inside it.
