Kenyan Space Startup Satlyt Raises $8M to Run AI on Satellites

Satlyt, led by founder and CEO Rama Afullo, raised an $8M seed round led by non sibi ventures to run AI apps on satellite computers already in orbit.

Four members of the Satlyt team posing in a room with a stone wall, with the Afrotech Horizon logo and a Startups and Funding label

Satlyt, a space software startup with offices in Nairobi and Sunnyvale, California, has raised an $8 million seed round to run artificial intelligence applications on satellites that are already in orbit. The round was led by US firm non sibi ventures, Space in Africa reported on October 1, citing the company's announcement.

The raise is a sizeable seed cheque for a deep-tech company with roots in Kenya. It also puts Satlyt in a growing market: squeezing more work out of the computers that satellites already carry.

Who backed Satlyt

Alongside non sibi ventures, the round drew a long list of investors, according to Space in Africa and Tech-ish:

  • TLCOM Capital
  • Antler
  • Slauson & Co.
  • Launch Africa Ventures
  • Enza Capital
  • Askya, Demos, BAG Collective and Gaingels
  • Axian Investment

Several of those names, including TLCOM, Enza and Launch Africa, are regular backers of African startups. Their presence next to US-based space investors suggests Satlyt is being pitched as both a global space company and an African tech story.

What the money is for

The company has not published a detailed spending breakdown. Its stated aim is to expand its software platform and sign more satellite operators and application developers.

What Satlyt actually does

Most satellites carry onboard computers that run narrow, fixed tasks. Satlyt builds software that lets operators run new applications on that existing hardware, including AI models, without launching new spacecraft.

Founder and chief executive Rama Afullo described the idea plainly: "Running applications on existing spacecraft gives operators a way to do more with the computing hardware they already have."

The practical payoff is bandwidth. Satellites collect far more data than they can send to the ground. If a model in orbit can filter, compress or analyse that data first, only the useful part needs to be transmitted.

The Gemma test

Satlyt's work features in a Google DeepMind case study on running Gemma, Google's family of open models, aboard a satellite. According to a roundup by FollowICT, Gemma 3 1B cut the volume of diagnostic data that needed to be sent to the ground by about 64%.

That figure comes from a specific diagnostic task. It should not be read as a general compression rate for every kind of satellite data.

The Transporter-18 mission

Satlyt's software was set to fly on SpaceX's Transporter-18 rideshare mission, which was targeted for October 1. The payload included research software developed with NASA Glenn Research Center under a Small Business Technology Transfer (STTR) award with the University of Houston, plus a commercial imagery application.

Two milestones matter here and they are separate. Launch is the first. Successful execution of Satlyt's software in orbit is the second. Public reports so far confirm the plan for the mission but have not confirmed results from the software running onboard. We will update readers when the company publishes them.

African partnerships

Satlyt has signed memoranda of understanding with the Kenya Space Agency and with GGPEN, the office that manages Angola's national space programme. MoUs are statements of intent rather than contracts, but they give the startup a route to work with two government space programmes on the continent.

For African agencies, software that adds capability to existing satellites is attractive. Building and launching new spacecraft is expensive, and many national programmes operate with tight budgets.

Why investors are interested

Kent Lucas, managing partner at non sibi ventures, framed the bet around software rather than hardware: "We see a vast opportunity for software to make spacecraft computing more useful to operators and application developers."

The round comes during an uneven year for African startup funding, where even the headline totals depend on who is counting, as we explained in our look at why African startup funding figures disagree. A deep-tech seed round of this size stands out against that backdrop.

What to watch

  • Confirmation that Satlyt's software ran successfully on Transporter-18 payloads
  • Whether the Kenya Space Agency and GGPEN agreements turn into paid projects
  • New operator customers outside the NASA research track

If the onboard results hold up, Satlyt will have shown that a team with roots in Nairobi can compete in one of the more specialised corners of the space industry. The next few months of mission data will decide how strong that case is.

Get the next one by email