Skip to content

AI operations and governance

Enterprise

Production AI needs an operating model, not only a provider key. Kumiko Enterprise keeps the controls close to the tenant, policy, handler, and audit boundaries of the application.

  1. Configure a provider and its secret for the tenant.
  2. Define the prompt and step defaults in code.
  3. Tune tenant policy or prompt revisions without changing the pipeline structure.
  4. Evaluate with mock responses and golden fixtures before activation.
  5. Activate or roll back the revision explicitly.
  6. Inspect provenance, token usage, provider, model, and prompt revision.
  7. Review executed writes in the normal audit path.
ControlWhere to start
Provider and model selectionai-foundation and provider reference
Prompt history and rollbackprompt-store
Draft, activate, and rollbackai-pipeline
Token and transcription budgetsThe Assistant and cap-counter
Recorded run facts and costProvenance and cost control
Launch evidenceProduction AI governance

Provenance can record identifiers and accounting facts such as provider, model, prompt revision, and token usage without storing the resolved prompt or model response. The in-memory step result still needs the application’s data classification before it is persisted or logged.

Assistant conversations and tool calls are application data. Apply the same field access, retention, export, erasure, secret handling, and backup rules as for other tenant and user data.

You can implement a smaller policy entity, evaluation runner, or audit report with public Kumiko features. Enterprise packages the revision model, activation commands, golden fixtures, dry runs, caps, and provenance conventions so teams do not have to assemble those controls repeatedly.