README.md

metadata

title: README
emoji: 🛰️
colorFrom: gray
colorTo: blue
sdk: static
pinned: false

Constellation Space

Constellation Space builds the open control plane for satellite and ground-station fleets: physics-grounded models that forecast link quality, traffic demand, and failures before they happen. Here we share models, datasets, and benchmarks so every operator, researcher, and student can bring frontier intelligence to space operations.

What we publish

• Models: forecasters for space operations, starting with link SNR (signal-to-noise ratio, 1 to 5 minutes ahead) and ground-station traffic demand. Each model card states its method, fidelity tier, inputs, and known limits.
• Datasets: telemetry, topology, and labeled outcomes from our demo fleet, with every vector tagged with its reference frame and time system.
• Benchmarks: prediction-versus-actual evaluations, so you can see how a model performs on real passes before you rely on it.

Use them in production

The same models run behind the Constellation API, so you can go from a notebook to live forecasts on your own fleet:

curl -H "x-api-key: $CONSTELLATION_API_KEY" https://api.constellation.space/topology

• Platform: platform.constellation.space. Create a key and load a demo fleet in minutes.
• Docs: platform.constellation.space/docs
• Agents: connect Claude, Cursor, or any MCP client to your fleet through our MCP server.

For students and researchers

Developer-plan allowances are free with a school email. See Constellation for Education.

Contact

constellation.space · X @constspace · LinkedIn · contact@constellation.space