Rama Afullo, a visionary Kenyan-American engineer, had a prescient understanding of the burgeoning potential for orbital data centers long before they became a focal point for aerospace giants. His journey through the high-stakes worlds of Google’s cloud computing division and SpaceX’s Starlink communications network in 2024 solidified his conviction that a massive, untapped opportunity lay in bringing advanced computing capabilities directly to space. This belief, initially met with skepticism at his former employers, has now culminated in the successful launch of Satlyt, a company poised to revolutionize in-orbit data processing with an $8 million seed round.
Afullo’s path was not without its early challenges. "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," he recounted to TechCrunch. Undeterred by these initial rejections, Afullo’s entrepreneurial spirit led him to depart SpaceX and co-found Satlyt. With dual headquarters established in Sunnyvale, California, and Nairobi, Kenya, Satlyt is dedicated to developing sophisticated software that enables multiple satellites to collaboratively undertake complex computing tasks directly in orbit. The recently secured $8 million seed round, exclusively revealed by Afullo to TechCrunch, underscores the growing investor confidence in his audacious vision.
The Shifting Landscape of Space Computing: A Timely Bet
Satlyt’s emergence arrives at a particularly opportune moment within the aerospace industry. The very giants that initially dismissed Afullo’s proposals have since pivoted significantly. SpaceX, recognizing the strategic imperative, has now fully embraced the concept of orbital data centers. Concurrently, Google is making substantial strides with its own space data center initiative, Project Suncatcher. In a significant validation of Satlyt’s trajectory, the company’s software is slated for launch this Thursday on a SpaceX rocket, sharing the payload bay with the inaugural prototype for Google’s Project Suncatcher. This concurrent launch not only highlights the rapid maturation of the in-orbit computing sector but also positions Satlyt as a critical, foundational player in this evolving ecosystem.
Unlike many of its contemporaries, including SpaceX, Google, and specialized startups such as Starcloud and Cowboy Space Company, Satlyt has strategically chosen not to engage in the capital-intensive and complex endeavor of building its own spacecraft. Instead, Afullo’s team is laser-focused on developing the advanced software necessary to operate sophisticated AI models directly on satellites. He draws a compelling analogy to established terrestrial tech giants: "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." This comparison positions Satlyt as a platform provider akin to VMware or Snowflake, companies whose innovative platforms empower users to run intricate software applications without the burden of managing the underlying infrastructure. Satlyt’s ambition is to offer a ubiquitous software solution that seamlessly operates across a diverse array of satellites manufactured by various companies, fostering an open and interoperable space computing environment.
The Imperative for In-Orbit Processing: Addressing Terrestrial Bottlenecks
The current paradigm for satellite operations is heavily reliant on ground-based flight controllers, who are responsible for resolving any issues that arise. This dependency often necessitates the costly and time-consuming process of "downlinking" data back to Earth. The expense of transmitting vast quantities of raw data, coupled with inherent latency issues, represents a significant operational and financial burden for satellite operators. The advent of robust in-orbit processing capabilities, powered by AI, promises to fundamentally alter this dynamic. By enabling satellites to make autonomous decisions and process sensor readings directly in space, operators can achieve substantial cost savings, both in resolving anomalies and in streamlining data analysis.
Satlyt has already demonstrated the tangible benefits of its approach through successful missions. Earlier this year, the company made headlines by deploying Google DeepMind’s Gemma AI model aboard a spacecraft operated by Momentus. This pioneering deployment resulted in a remarkable reduction of over 60% in the size of transmissions related to onboard software errors. Afullo estimates that this kind of operational efficiency can translate into savings of hundreds of thousands of dollars per satellite annually, a compelling economic argument for broader adoption.
A Pivotal Launch: Expanding Capabilities and Validating Vision
The upcoming launch marks a critical milestone for Satlyt, as its software will be integrated into a spacecraft developed by TakeMe2Space, an innovative Indian startup specializing in satellite computing hardware. This mission is designed to serve a tripartite customer base, showcasing the versatility and robustness of Satlyt’s platform. NASA, a key partner, is funding the company to rigorously test protocols for cloud computing in the demanding space environment. Stellerian, another startup focused on space surveillance and the tracking of orbital objects, aims to evaluate the efficiency of in-orbit image processing workloads. Finally, TakeMe2Space itself will leverage the mission to demonstrate its satellite’s capacity to host and run third-party software, effectively positioning it as a foundational node in a nascent orbital cloud network.
This collaborative approach highlights a broader trend in the space industry: the move towards modularity, interoperability, and shared infrastructure. As the number of active satellites continues to proliferate—with estimates suggesting tens of thousands more will be launched in the coming decade, driven by constellations like Starlink, OneWeb, and Kuiper—the demand for efficient data management and processing will only intensify. Current projections indicate that the global satellite services market, a key segment of the broader space economy, is set to exceed $300 billion by 2030, with data processing and analytics forming an increasingly critical component of this growth. Satlyt’s strategy directly addresses this escalating demand by providing the software backbone for a distributed, intelligent network of space assets.
Investment and Strategic Foresight: Beyond the "Highest-Octane Vision"
Satlyt’s seed round was spearheaded by Houston-based Non Sibi Ventures, a firm with a deep understanding of the space sector. Partner Bernard Harris, a former NASA astronaut with over two decades of experience, played a crucial role in validating the firm’s conviction in Satlyt’s novel, space-based business model. Non Sibi Ventures’ investment strategy reflects a pragmatic understanding of the evolving space economy.
As Non Sibi partner Kent Lucas elucidated to TechCrunch, "We don’t need data centers in space for Rama to be wildly successful, right? It can just be driven by the number of satellites going up." This statement underscores a key differentiator for Satlyt: its business model is not solely predicated on the "highest-octane vision" of massive, dedicated space data centers, which face considerable economic uncertainties and are bottlenecked by launch availability. Instead, Satlyt’s success is intrinsically linked to the inexorable growth in satellite deployments. Every new satellite launched, particularly those equipped with advanced processors, represents a potential client for Satlyt’s software, transforming the satellite into a revenue-generating managed service. Afullo articulates this vision succinctly: "The way to think about it is that we turn your satellite into a revenue-generating managed service."
The Broader Landscape of Space Computing: Opportunities and Challenges
While the full realization of high-powered GPU farms in orbit is still in its nascent stages due to challenges related to radiation hardening, power consumption, and thermal management, the immediate opportunities lie precisely in the types of workloads Satlyt is pioneering. These include edge computing for anomaly detection, real-time data compression, and initial processing of sensor data. However, Satlyt’s ambition extends beyond these initial applications. The company’s next major project aims to demonstrate the creation of a shared computing system—a true "cloud"—spanning two distinct satellites, a feat it expects to attempt next year. Should this prove successful, it would pave the way for Satlyt to establish itself as a third-party provider of a genuine compute cloud in orbit, offering scalable and flexible computing resources to spacecraft builders and operators globally.
The broader implications of this shift are profound. The ability to process data at the source, rather than exclusively on Earth, opens up new possibilities for real-time Earth observation, enhanced navigation systems, sophisticated space situational awareness, and even in-orbit manufacturing and assembly. This decentralization of computing power could democratize access to advanced space capabilities, fostering innovation across a wider spectrum of industries.
However, the path to widespread orbital computing is not without its hurdles. Beyond the technical challenges of operating sophisticated hardware in the harsh radiation environment of space, there are significant considerations around software updates, cybersecurity in a distributed network, and the regulatory frameworks governing data processing in international space. Furthermore, the sheer volume of data generated by future satellite constellations will demand ever more robust and efficient processing solutions, pushing the boundaries of current technology.
Future Trajectory and Ambitious Goals
Afullo harbors ambitious goals for Satlyt, envisioning the company’s software being live on 20% of all operational satellites by the end of the decade. This projection is underpinned by his strong belief that by 2030, virtually every spacecraft builder will be incorporating GPUs and similar advanced processors into their satellite designs. "If you’re putting up a satellite without putting up a GPU on it, at the very least, you’re doing yourself a disservice, right?" he posits, underscoring the growing imperative for onboard intelligence.
The integration of GPUs and AI into satellite architecture is not merely an optional upgrade; it is becoming a fundamental requirement for maximizing the utility and economic viability of space assets. As the cost of launching payloads decreases and the capabilities of onboard processing units increase, the business case for performing compute-intensive tasks in orbit becomes overwhelmingly compelling. Satlyt, by focusing on the software layer, is positioning itself as the critical enabler for this paradigm shift, allowing hardware manufacturers and satellite operators to leverage these advanced capabilities without developing proprietary software stacks.
Satlyt’s participation in TechCrunch’s Battlefield competition at Disrupt in downtown San Francisco later this year further highlights its visibility and potential within the startup ecosystem. As the space economy continues its rapid expansion, driven by both governmental and private sector investments, companies like Satlyt are poised to define the next frontier of innovation, transforming satellites from mere data collectors into intelligent, autonomous, and revenue-generating nodes in an interconnected orbital network. The successful seed round and imminent launch represent not just a victory for Satlyt, but a significant leap forward for the entire orbital computing industry, signaling a new era where the cloud extends far beyond Earth’s atmosphere.
