Execute workloads across Earth and orbit.
Kubernetes doesn't understand orbital mechanics. PyTorch doesn't adapt to solar eclipses. TensorFlow doesn't expect bit flips. The runtime is the layer that does - it plans your workload around the orbit, runs it on the satellite, and keeps it alive through eclipse, power limits, and radiation.
Plan it. Run it on-orbit. Keep it alive.
$ curl https://api.rotastellar.com/v1/plan \ -H "Content-Type: application/json" \ -d '{ "satellite_id": "25544", "preset_id": "onboard-ml-inference" }' # CAE returns a constraint-aware plan { "satellite": { "name": "ISS (ZARYA)", "altitude_km": 408 }, "preset": { "id": "onboard-ml-inference", "steps": 4 }, "orbital_environment": { "eclipse_fraction": 0.35 }, "plan": { "total_duration_s": 2700, "windows_used": 8, "status": "scheduled" }, "error_budget": { "delivery_confidence": 0.98 } }
Three primitives. New foundations.
Placement that understands orbits
Decide where a workload runs, when, and at what fidelity - from orbital mechanics, energy availability, and link windows.
Explore →Inference that bends, not breaks
Inference and training that adapt to available energy and thermal headroom instead of failing when the budget tightens.
Explore →When bit flips are normal
Fault-tolerant ML execution for radiation environments, where single-event upsets are nominal, not exceptional.
Explore →Stateless orbital computation
Propagate positions, detect eclipses, and compute ground passes - the digital twin the other three are built on.
Explore →Put your first workload in orbit.
Tell us what you want to run and the satellite you want to run it on. We'll come back with a feasible plan - or an honest reason it can't be done yet.