Rama Afullo and his team at Satlyt should be celebrating their successful test earlier this month—and they are. Working with the University of Houston on a NASA-funded research project after 13 months of development, the company demonstrated that its software can spread computing tasks among orbital computers based on communication and available compute. It's a milestone worth acknowledging. But Afullo's own candid assessment reveals something more important: this was merely one building block toward satellites sharing computing work. The real test comes next.
The test itself was appropriately scoped. Satlyt created a simulated network with four computers, four users, and eight services. The software assessed how to assign and run tasks, accounting for compute and communication availability. One physical computer aboard a spacecraft ran the experiment, then returned results to Earth. The software successfully generated plans on hardware in orbit—proving the concept works. But here's the catch that Afullo himself highlighted: more testing will be done to see how those computers transfer information amongst each other in practice when multiple physical computers are used. The simulation and the messy reality of space are not the same thing.
What makes this moment particularly interesting is how honest Afullo was about the obstacles still ahead. Yes, the software development, onboard execution, and return of outputs succeeded. But Satlyt also discovered a critical bug: it was time-referencing delays incorrectly, stamping arrival time rather than elapsed delay. The company needs usable communication links, permission from participating operators, secure transfers, resource scheduling, and recovery protocols when a connection becomes unavailable. These aren't minor details. They're the infrastructure that separates a working prototype from a deployable system.
The broader mission is ambitious and compelling. Founded in 2024 by Afullo, who previously worked at Google and SpaceX, Satlyt is trying to turn spacecraft into virtual data centers where AI and other applications run onboard, processing data closer to where it's collected. The vision—enabling wildfire detection software to transmit relevant information faster, supporting 4K video streaming from the Moon, allowing developers to build space applications without rebuilding infrastructure for every mission—matters for humanity's relationship with space.
Still, it's worth noting that while this test used NASA-funded research, NASA has not yet selected Satlyt for any particular mission or chosen to deploy its technology. The company was in the top five of TechCrunch Disrupt Battlefield this year, which speaks to investor enthusiasm. But enthusiasm isn't deployment. Afullo is correct that the next technical steps are broader validation and integration, followed by demonstrations of computing work shared across separate spacecraft. Until those happen at scale and in real orbital conditions, this remains a promising proof of concept rather than a proven solution. That distinction matters when you're building the infrastructure for space.
Satlyt has the right founder, clear technical direction, and demonstrated proof of concept. What it needs now is the unglamorous work of validation, integration, and real-world testing—the stuff that turns great ideas into indispensable infrastructure. The GOOOO moment has passed. Now comes the patient, methodical work that will determine whether this company actually transforms how satellites operate.