We do not sell seats; we take on partners. Each profile below is a real class of collaborator, university groups, observatories, and climate organizations, described in anonymized form to protect ongoing work and grant terms.
Three classes of partner, each leaning on a different face of the runtime, and all sharing the same provenance and budget model underneath.
A simulation-heavy theory group spun out of a national lab. They live in Nova-Sim, running radiative-transfer ensembles that have to restart bit-for-bit after preemption.
A wide-field observatory that moved 8 PB of legacy frames into the Nova-Obs asset store. They calibrate at the telescope and serve the same objects to downstream theory groups by hash.
A multi-institution climate body running downscaling ensembles. Nova-Loop plans their runs against a fixed annual grant and stops when convergence, not the calendar, says so.
A facility that hosts Constellation for dozens of small groups at once. The per-grant scheduler lets unrelated labs compete fairly for the same nodes without manual quotas.
An instrument-building team that pairs Nova-Obs ingestion with Nova-Sim forward models to validate a detector before first light, all under one provenance graph.
One principal investigator, one postdoc, one hard problem. They use the full runtime on a modest academic tier and cite Constellation directly in every paper.
Every partner started the same way: a hard problem, a scoped engagement, and a runtime that grew to fit it. Here is the shape that usually takes.
We sit with your PIs, look at the real workload, and decide whether Constellation is even the right tool.
Budgets are bound to grants, data lands in the asset store, and the first runs go through the scheduler.
More PIs, more grants, more modules. Provenance keeps the whole expanding picture reproducible.
Results ship with signed cost-and-lineage exports. Many partners co-author the methods paper with us.
If your group has a problem too big for a dashboard and too important to get wrong, we would like to hear about it. Partnerships begin with a conversation, not a contract.