Skip to content

Deployment and models

Option Details
Deploy your way Managed SaaS or self-hosted in your environment.
Your model Local models or an approved provider — switchable without changing workflows.
Full provenance Every AI response records provider, model and configuration.

Every model call goes through one model gateway, and every AI response records which provider, model and configuration produced it. Responses calculated without a model are marked as deterministic and are never presented as AI-generated. Provenance never contains API keys or other secrets.

Logs, code and incident text are redacted before any model sees them: personal data, exact customer identifiers and credentials are removed. Provenant does not retrain a model on your data.