AI jobs at space companies: fewer than you'd think, and almost all senior
We counted every open posting at 11 space companies. About 2.6 percent had AI or machine learning in the title. Where they are and how to get one.
On Oct. 21 we pulled every open posting from the public job boards of 11 space companies: SpaceX, Rocket Lab, Relativity Space, Vast, Astranis, Planet, Muon Space, Varda, Inversion, Northwood and ispace. That came to 4,711 roles. Just 122 had AI, machine learning, autonomy, perception, computer vision or data science in the title, about 2.6 percent. SpaceX accounted for 103 of them.
And 119 of the 122 were experienced roles by our title-based sort. We found two entry-level titles (including an “AI Software Engineer I” at Relativity) and one internship, at Inversion.
So the honest answer to “how do I get an AI job at a space company” is: mostly, by first being good at AI somewhere else. The space companies hiring ML people want them to point existing skills at a narrow problem. Look at the titles. Rocket Lab is hiring machine learning engineers for EO/IR image analysis. Muon Space wants remote sensing data scientists and computer vision engineers. Planet hires data scientists. Relativity lists an “AI/ML Scientist, Planetary Science.” These are Earth observation, imagery and manufacturing problems more often than flight autonomy.
That suggests two routes in.
The data route. Satellite imagery is the biggest AI market inside space, and it’s the one with public data. A portfolio project that does something useful with open imagery (change detection, cloud masking, crop or ship counting) speaks directly to Planet, Muon and Rocket Lab’s imaging teams. Our portfolio guide has more on scoping one.
The autonomy route. Spacecraft and rovers need navigation and perception, and the skills overlap almost completely with drones and self-driving cars. MIT’s free Visual Navigation for Autonomous Vehicles course covers the stack: ROS, trajectory following, feature detection, localization and SLAM. Our board’s robotics companies (Skydio, Zipline, Waymo, Nuro and the humanoid makers) hire for exactly this at far higher volume, and with fewer export-control limits; see our ITAR guide.
A caveat on method: we classify by job title only, using the same script that labels our jobs board. Plenty of software roles involve ML without saying so, so treat 2.6 percent as a floor.
Our view: if you want AI and space, get hired at an AI or robotics company first and move over with a track record. The space side is hiring specialists, not trainees.
What would change our mind: a space company posting a batch of entry-level ML roles or a dedicated AI internship class. We’d flag it in the briefing’s Opportunities section the week it happens.
Not affiliated with NASA, SpaceX or any agency or company mentioned. Not investment advice. Disclaimer.