Home · Technology · Oct 1 archive
Satlyt Raises $8M for AI on Satellites
Confirmed
In Short: Satlyt, a startup founded by former SpaceX and Google employee Rama Afullo, has raised $8 million to develop AI software for satellites.
Satlyt, a startup founded by former SpaceX and Google employee Rama Afullo, has raised $8 million in seed funding to develop AI software for satellites. The company aims to enable satellites to share large computing tasks in space, reducing the need for extensive data transmission.
Satlyt’s software, which will launch on a SpaceX rocket alongside Google’s Project Suncatcher prototype, is designed to operate AI models on satellites without the need for specialized hardware. This approach is similar to how companies like VMware and Snowflake provide platforms for running complex software.
According to Afullo, Satlyt’s software has already demonstrated its ability to process imagery, generate encrypted downlink-ready outputs, and diagnose software faults in space. The company recently deployed Google DeepMind’s Gemma AI in space, significantly cutting transmission costs.
Unlike competitors such as SpaceX, Google, and Starcloud, Satlyt does not manufacture spacecraft. Instead, it focuses on building a horizontal operating system that can run across diverse satellite platforms.
Satlyt’s software allows satellites to process data in orbit and downlink only the most necessary data outputs, addressing the challenge of efficiently transmitting large datasets back to Earth.
In November, Satlyt won a NASA STTR contract worth $4 million to design delay-tolerant edge protocols in space, further advancing its mission to enable in-orbit AI processing.
With the new capital, Satlyt plans to grow its network and increase its in-orbit capabilities. Afullo envisions Satlyt’s software proliferating across non-Starlink satellites, aiming to onboard 30% of all non-Starlink satellites by the end of the decade.
Satlyt positions its software as an open ecosystem, differentiating itself from competitors who act like the iPhone by focusing on building the underlying architecture for in-orbit AI processing.
While investors pour billions into orbital data center concepts, Satlyt’s virtual ODC network is already flying above us and growing rapidly.
Satlyt’s software has flown on two demonstration missions, reducing the financial delay of transmitting massive raw datasets back to Earth.
What's confirmed
- Satlyt, a startup founded by former SpaceX and Google employee Rama Afullo, has raised $8 million in seed funding to develop AI software for satellites. The company aims to enable satellites to share large computing tasks in space, reducing the need for extensive data transmission.
- Satlyt’s software, which will launch on a SpaceX rocket alongside Google’s Project Suncatcher prototype, is designed to operate AI models on satellites without the need for specialized hardware. This approach is similar to how companies like VMware and Snowflake provide platforms for running complex software.
- According to Afullo, Satlyt’s software has already demonstrated its ability to process imagery, generate encrypted downlink-ready outputs, and diagnose software faults in space. The company recently deployed Google DeepMind’s Gemma AI in space, significantly cutting transmission costs.
- Unlike competitors such as SpaceX, Google, and Starcloud, Satlyt does not manufacture spacecraft. Instead, it focuses on building a horizontal operating system that can run across diverse satellite platforms.
- Satlyt’s software allows satellites to process data in orbit and downlink only the most necessary data outputs, addressing the challenge of efficiently transmitting large datasets back to Earth.
- In November, Satlyt won a NASA STTR contract worth $4 million to design delay-tolerant edge protocols in space, further advancing its mission to enable in-orbit AI processing.
- With the new capital, Satlyt plans to grow its network and increase its in-orbit capabilities. Afullo envisions Satlyt’s software proliferating across non-Starlink satellites, aiming to onboard 30% of all non-Starlink satellites by the end of the decade.
- Satlyt positions its software as an open ecosystem, differentiating itself from competitors who act like the iPhone by focusing on building the underlying architecture for in-orbit AI processing.
- While investors pour billions into orbital data center concepts, Satlyt’s virtual ODC network is already flying above us and growing rapidly.
- Satlyt’s software has flown on two demonstration missions, reducing the financial delay of transmitting massive raw datasets back to Earth.
What's still developing
- For now, the most likely opportunity lies in the kinds of workloads Satlyt is working on.
- It brings Satlyt’s total funds to “well past $10M,” according to CEO Rama Afullo.
- Afullo told Payload that the long-term vision is to “proliferate across basically every non-SpaceX Starlink satellite,” with the goal of onboarding 30% of all non-Starlink satellites flying in orbit by the end of the decade.
- Meet Satlyt: Satlyt was founded in 2024 by Afullo, an alum of SpaceX and Google who spent years trying to convince both companies to invest in in-space edge computing opportunities.