Beyond the Horizon: How Satlyt is Building the Software Backbone for Orbital Cloud Computing

By Tim Fernholz

Main Facts

The commercial space race is hurtling toward a new frontier, and the bottleneck is no longer just getting into orbit—it is what happens to the data once you get there. Enter Satlyt, an enterprise software startup bridging the gap between artificial intelligence and space infrastructure. Co-founded by Kenyan-American engineer Rama Afullo, Satlyt has officially emerged from stealth mode with an $8 million seed funding round led by Houston-based Non Sibi Ventures.

Unlike heavyweights such as SpaceX, Google, or ambitious hardware startups like Starcloud and Cowboy Space Company, Satlyt is not building rockets or physical spacecraft. Instead, the Sunnyvale- and Nairobi-based startup is positioning itself as the "Android" of space: a horizontally integrated, open-ecosystem software provider. Its mission is to enable AI models and distributed computing workloads to run seamlessly across disparate satellites, transforming isolated spacecraft into a unified, revenue-generating, managed cloud service.

As the industry prepares to launch Satlyt’s software on an upcoming SpaceX Falcon 9 rocket—flying alongside Google’s Project Suncatcher prototype—the company is rewriting the economic rules of modern aerospace engineering. By processing data on the edge and bypassing expensive ground-station downlinks, Satlyt aims to save satellite operators hundreds of thousands of dollars per vehicle annually while laying the groundwork for the world’s first true multi-satellite orbital computing cloud.

Chronology: From Big Tech Rejection to Orbital Innovation

The genesis of Satlyt is a classic tale of visionary persistence in the face of corporate caution.

The Tech Giant Stints (2024)

Rama Afullo spent his formative years cutting his teeth in the high-stakes environments of Silicon Valley. He worked as an engineer in Google’s premier cloud computing division before transitioning to a brief, high-intensity stint at SpaceX’s Starlink communications network in 2024. Immersed in the rapidly scaling satellite internet ecosystem, Afullo recognized an impending paradox: while thousands of satellites were being launched into low Earth orbit (LEO), they remained functionally isolated, lacking the software infrastructure required to share computing resources, process heavy workloads locally, or run advanced AI models.

"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 recalls.

Stepping Out and Securing Backing

Unwilling to let the idea gather dust, Afullo departed SpaceX to build the solution himself. Alongside co-founders, he established Satlyt with a dual-headquarter model spanning Sunnyvale, California, and Nairobi, Kenya, bridging Silicon Valley venture capital with emerging global tech talent.

The company recently finalized an $8 million seed round. The investment was led by Non Sibi Ventures, whose partner Bernard Harris—a veteran NASA astronaut with over 20 years of experience—recognized the profound commercial viability of a software-first approach to space infrastructure.

Early Demonstrations (Early 2025)

Before seeking massive institutional validation, Satlyt put its code to the ultimate test. Earlier this year, the company successfully deployed Google DeepMind’s Gemma AI model onboard an active spacecraft operated by aerospace logistics firm Momentus. The onboard AI deployment successfully cut the data transmission size of software error logs by more than 60%. This reduction in downlink data volume eliminated significant bandwidth costs, proving that edge computing in space could deliver immediate, measurable ROI.

The Milestone Falcon 9 Launch

The timing of Satlyt’s market entry aligns with a broader industry awakening. SpaceX has since shifted its strategic focus toward orbital data centers, and Satlyt’s software is scheduled to launch on an upcoming SpaceX mission. Notably, the payload shares space with the inaugural prototype for Google’s own orbital data center initiative, Project Suncatcher.

Supporting Data and Technical Architecture

Satlyt’s business model deliberately sidesteps the crushing capital expenditures and supply-chain bottlenecks associated with manufacturing and launching hardware. Instead, the company draws comparisons to foundational terrestrial software giants like VMware and Snowflake—platforms that allow enterprises to leverage complex software ecosystems without needing to manage the underlying physical hardware.

The Economic Burden of "Downlinking"

Traditionally, satellites act as glorified sensors. They capture imagery, telemetry, or scientific data and beam it back down to ground stations via expensive radio frequency links. This process, known as downlinking, is notoriously slow, subject to atmospheric interference, and heavily constrained by limited ground-station visibility windows.

If a satellite encounters an anomaly or anomaly-like sensor reading, flight controllers on Earth must manually review the telemetry, diagnose the problem, and upload a patch. This latency can cost days of operational downtime.

By utilizing Satlyt’s software architecture:

• Edge Processing: AI models run directly on the spacecraft’s onboard processors.
• Bandwidth Preservation: Instead of transmitting raw gigabytes of data back to Earth, the satellite processes the information locally and transmits only lightweight, actionable insights.
• Cost Efficiency: The Momentus mission demonstrated that optimizing error-log transmissions alone saves upwards of hundreds of thousands of dollars per satellite each year.

The Upcoming Multi-Customer Mission

The imminent SpaceX launch will deploy Satlyt’s software stack onto a satellite built by TakeMe2Space, an innovative Indian startup specializing in orbital computing hardware. This mission serves as a critical proof-of-concept for three distinct stakeholders:

1. NASA: Funding the flight to rigorously test protocols for cloud computing standards in space.
2. Stellerian: A space-surveillance startup tracking orbital debris and space traffic, utilizing the flight to test advanced on-orbit image processing workloads.
3. TakeMe2Space: Utilizing the mission to demonstrate its hardware’s capability to natively host third-party software ecosystems.

Toward a Multi-Satellite Distributed Cloud

While high-powered Graphics Processing Units (GPUs) remain relatively rare in orbit today, Satlyt is already looking ahead to its next evolutionary phase. Next year, the company plans to execute its most ambitious test yet: creating a shared computing network—a localized cloud—spanning two different satellites.

If successful, Satlyt will transition from a software optimization tool into a true third-party cloud provider for spacecraft builders, turning individual satellites into revenue-generating, managed infrastructure nodes.

Official Responses and Industry Perspectives

The rapid rise of Satlyt has caught the attention of veteran aerospace investors who understand the friction points of traditional mission architectures.

Non Sibi Ventures partner Kent Lucas noted that Satlyt’s appeal lies precisely in its independence from the speculative, high-risk economic models governing science-fiction-style space data centers.

"We don’t need data centers in space for Rama to be wildly successful, right?" Lucas told TechCrunch. "It can just be driven by the number of satellites going up."

This pragmatism resonates with the current state of the market. While massive orbital data centers face profound logistical hurdles—including the monumental task of radiating away thermal heat in the vacuum of space and dealing with current launch capacity bottlenecks—the sheer volume of operational LEO satellites is surging exponentially.

Rama Afullo views the integration of advanced processing power as an inevitable evolutionary baseline for every aerospace manufacturer. His stated ambition is to have Satlyt’s software running on 20% of all global satellites by the end of the decade.

"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?" Afullo asserts. "We turn your satellite into a revenue-generating managed service."

Implications: Reshaping the Space Economy

Satlyt’s emergence signals a mature transition in the commercial space industry. For decades, the space economy was defined by hardware milestones: heavier rockets, larger constellations, and more sensitive optical sensors. Software was historically treated as proprietary, closed-source firmware deeply coupled with specific hardware components.

By decoupling software from hardware—much like the smartphone industry did two decades ago—Satlyt is unlocking several profound systemic shifts:

1. Democratization of Space Capabilities: Smaller startups that cannot afford custom software development teams can now plug into Satlyt’s ecosystem to run sophisticated AI workloads, leveling the playing field against legacy aerospace primes.
2. The Rise of "Space-as-a-Service": By turning idle orbital compute cycles into a distributed cloud, satellite operators can monetize excess capacity, shifting their assets from depreciating capital expenses into recurring revenue streams.
3. Autonomous Orbital Management: As low Earth orbit becomes increasingly crowded with tens of thousands of active satellites and debris pieces, relying on ground-based human intervention for collision avoidance and system diagnostics is no longer viable. Edge-native AI platforms like Satlyt will become essential safety infrastructure for the sustainable management of orbital space.

As Satlyt prepares to showcase its technology at prominent industry showcases like TechCrunch Disrupt in San Francisco, the message to the aerospace community is clear: the future of space is not just about building better metal; it is about writing smarter code.