Satellite software company Satlyt has raised $8 million in seed funding to put AI processing aboard spacecraft, in a round that includes several of Africa’s more active early-stage investors.
The round was led by non sibi ventures, with participation from TLCOM, Antler, Slauson & Co., Launch Africa Ventures, Enza Capital, Askya Investment Partners, Demos, BAG Collective, Gaingels, Axian Investment and existing backers.
Capital will fund product development, hiring and wider deployment of Satlyt’s software on third-party satellites.
The problem
Satellites capture imagery, measurements and system data continuously. Communications capacity to get that data back to Earth is limited, which delays analysis.
Satlyt’s approach is to process in orbit instead — moving computing tasks onto the spacecraft so operators receive usable information faster, and allowing developers to deploy applications directly on satellite computing systems.
That is edge computing taken to its furthest point. The reasoning is the same as running AI at a base station rather than routing it to a central data centre: process where the data is, when the link back is the constraint.
Already in orbit
Satlyt has software running on satellites now. An earlier deployment ran Google’s Gemma AI model aboard a spacecraft to analyse system logs, software errors and stack traces without relying on ground processing.
That is a working deployment rather than a demonstration, and it distinguishes this from most space-adjacent AI announcements.
Next up are two applications aboard a third-party spacecraft: research software tied to Satlyt’s NASA Glenn Small Business Technology Transfer work with the University of Houston, and a commercial imagery-processing application. No launch date was given.
The longer-term objective is a common software layer coordinating computing resources across satellites operated by different organisations — in effect, an orbital compute network.
Kenyan or American
Satlyt has US headquarters in Sunnyvale, California and African headquarters in Nairobi. Leadership is Kenyan-American, with significant engineering presence in Nairobi.
Which label applies depends on the measure. The investor base reflects the same split — TLCOM, Launch Africa, Enza, Axian and Askya are Africa-focused; non sibi, Slauson & Co. and Gaingels are not.
That is a familiar shape. Moove was founded in Lagos and now headquarters in Abu Dhabi. Terra Industries was founded in Nigeria and is opening offices in London, San Francisco and Washington. Satlyt differs in retaining engineering in Nairobi rather than relocating it — which is the part that matters for whether capability stays.
Why orbital compute, and why now
Every African AI infrastructure story of the past year has concerned ground-based capacity: Teraco’s 30MW JB5, Equinix’s phased Johannesburg build, Morocco’s 500MW Nexus AI Factory, the stalled Olkaria project in Kenya, Lesotho’s compute roadmap.
Satlyt sits outside that entirely. No grid connection, no water for cooling, no municipal approval. For a continent where power and water have become the binding constraints on data centre development — and the subject of a legal challenge in Cape Town — compute that runs in orbit is a different proposition.
It is also a small market. Satellite onboard computing serves operators, not enterprises, and $8 million is a seed round rather than infrastructure capital.
Askya’s first visible move
The round marks the first deployment iAfrica has seen from Askya Investment Partners, founded last year by Babacar Seck, the former Digital Africa chief executive who previously backed Moniepoint, Jumia, GoMyCode and Complete Farmer at Proparco.
Askya opened applications in September for a six-week programme for ten African AI startups, with a commitment of up to $200,000 to at least one participant. Seck’s stated thesis is backing companies with working products and real traction rather than funding at idea stage — which Satlyt, with software already in orbit, fits.
Satlyt did not name its founders, disclose the size of its Nairobi team, or identify customers.
