Satlyt, a startup based in Sunnyvale and Nairobi, has secured $8 million in a seed funding round led by Non Sibi Ventures. The company develops software designed to run AI models directly on satellites, aiming to optimize spacecraft operations and reduce the need for ground-based data processing.
Unlike competitors building their own hardware, Satlyt focuses on creating a platform for third-party satellites, similar to a cloud provider. The company has already completed two demonstration missions, including one that utilized Google DeepMind’s Gemma model to reduce transmission data regarding software errors by over 60%.
The upcoming mission features a collaboration with TakeMe2Space, NASA, and Stellerian to test cloud computing protocols and image processing. Satlyt plans to demonstrate a shared computing system spanning multiple satellites next year, with the goal of being present on 20% of all satellites by 2030.
