Retire old data without breaking what depends on it

Payne creates verified Shutdown Gates from DataHub evidence and invalidates approvals when monitored assumptions stop being true. It is for data platform and analytics engineers completing a dataset retirement with a deadline.

A Shutdown Gate is the durable record of the conditions, evidence, application work, human authority, and monitoring required to retire one dataset. Its approval is self-invalidating: it remains valid only while those recorded assumptions remain true.

Choose your path

• Try the guided simulation to inspect the complete Day 1 and Day 2 story in a hosted, fixture-backed experience. It is not current DataHub evidence.
• Run Payne locally to connect the local-live stack to its bundled DataHub scenario, PostgreSQL, worker, inference provider, sandbox, and re-ingestion path.

Before choosing, compare operating modes. Local-live is for development and evaluation; Payne does not claim a production deployment profile.

What Payne controls

DataHub owns current catalog facts. PostgreSQL owns Payne’s durable records. A configured model may investigate and interpret bounded evidence, but deterministic rules compute the gate outcome and an accountable human retains shutdown authority. A prepared or accepted patch is not applied evidence, and a successful mutation response is not verified until a fresh read matches.

Next, create your first Shutdown Gate or inspect the architecture.