Realtime AI News
Satlyt, founded by an ex-Google and SpaceX product manager, raises $8M to run AI on satellites
Satlyt, a satellite-computing startup founded by a former Google and SpaceX product manager, has raised $8 million to run AI directly on satellites. It positions itself as the Android of orbital computing, offering open software that runs across many companies' satellites in contrast to SpaceX's closed, all-in-one iPhone-style approach.

Satlyt, a satellite-computing startup founded by a former Google and SpaceX product manager, has raised $8 million to run AI directly on satellites. The funding is aimed at pushing artificial intelligence out of the data center and into orbit, where the data is actually generated.
The company's pitch is to become the Android of orbital computing: an open software layer designed to run across satellites built by many different companies, rather than being tied to a single vendor's hardware stack. That approach is a deliberate contrast with SpaceX's closed, all-in-one iPhone-style model.
In other words, the contest is not only about who can put a stronger model in orbit, but about who defines the software standard for running AI there. If open software is widely adopted, satellite operators would not have to rewrite an entire inference stack for every hardware generation, and the pace of iteration for compute and applications would shift accordingly.
The practical case for on-orbit AI comes down to latency and bandwidth. Doing inference locally on a satellite reduces reliance on downlink capacity to ground stations, letting remote sensing and monitoring workloads be judged at the moment the data is captured instead of queuing for transmission and processing.
Seen at industry level, this is a classic ecosystem fight. Closed, integrated systems often ship faster and deliver a more uniform experience early on, but an open layer that reaches developer scale can win a longer life in a market with diverse hardware.
$8 million is not a large sum in the space sector, but for an early-stage company trying to define a software layer it is more of a validation ticket: enough to move a prototype toward deployment and enough to persuade a first group of satellite operators to try it on real missions.
Two things are worth watching: whether Satlyt can turn 'open' from a slogan into usable interfaces and toolchains, and whether satellite manufacturers and operators publicly line up behind the approach. If they do, the fight over orbital computing's software layer will have only just begun.
Why it matters
If an open software layer is broadly adopted by satellite builders, the standardization battle for AI inference extends from the cloud into orbit, giving constellation operators far more freedom in hardware choices. If not, closed integrated systems may keep leading on delivery speed.
Nearby Updates
All10/01, 21:00
Legato launches AI hearing glasses starting at $999
Hearing tech startup Legato announced on Thursday, October 1, that its flagship AI hearing glasses are now available to buy, starting at $999 for adults with up to moderate hearing loss. Legato Frames embed the company's patented hearing-assistance tech in eyewear arms and use AI to separate voices from background noise instead of the directional microphones used by traditional hearing aids.
10/01, 21:00
Productive, Durable, Fungible: How NVIDIA AI Factories Maximize Return on Investment
Productive, Durable, Fungible: How NVIDIA AI Factories Maximize Return on Investment. AI factories are built by the megawatt, even by the gigawatt. Each megawatt factory costs roughly $60 million, and AI factory operators will only commit capital on that scale with a clear view of the return on investment. Three key things shape AI factory retu...
10/01, 07:43
Google releases Gemini 4 Argon, calling it its most powerful model yet
Google has released Gemini 4 Argon, describing it as its most powerful model yet and positioning it as a workhorse for coding and cybersecurity work. The TechCrunch report offers little beyond that framing, leaving parameters, context length, pricing, and availability undisclosed.
10/01, 05:07
Valor, Atreides and Sequoia back Flow Engineering at a $750M valuation
TechCrunch reports that Flow Engineering, a startup applying AI agents to hardware design, has raised a new round from Valor, Atreides and Sequoia at a $750 million valuation. Roelof Botha also joined as an angel investor and will sit on the company's board.