Contract API
When an AI agent asks “what is our churn rate?”, it should get the governed definition, owner, version, validation state, and approval context — not a stale warehouse artifact.
Start FreeThree churn definitions in the warehouse. The Finance version, the Marketing version, and a draft someone forgot to delete. Which one does the agent use?
The agent uses the wrong definition. It auto-triggers win-back campaigns targeting customers who are still active. Fast, confident, and wrong.
The agent queries ClariLayer and gets the governed contract context: current version, validation state, approval context, and what the metric is allowed to power.
An AI agent, BI tool, or internal app asks: “What is MRR?” The API returns the organization’s canonical definition.
The response includes the governed definition, SQL logic, business description, owner, approval status, validation date, tier, lifecycle state, and version.
Each API lookup is grounded in a specific metric version, with lifecycle state, approval context, and validation evidence attached.
Integration Surface
AI agents are dangerous when they treat every table, YAML file, and dashboard query as equally authoritative. The Contract API gives those systems a governed lookup before execution: which definition is current, who owns it, what policy tier it sits in, whether the release has validation and approval evidence attached, and which consumers should treat it as safe.
The API is downstream of the full lifecycle. Definitions come from AI-assisted Metric Studio authoring, proof comes from warehouse validation checks, and trust state comes from governed release workflows. Consumers can then use the Metrics Contract API documentation to understand the public response shape.
Agents can ask for the approved business definition before drafting SQL, choosing an automation path, or explaining a KPI to a user. The response gives the agent policy, lifecycle status, and evidence, not just a label.
Dashboards, notebooks, and operating tools can display owner, lifecycle status, version, validation recency, and approval context beside the metric so business users understand whether the number is suitable for the decision at hand.
API consumers receive a particular metric version and lifecycle state. If a definition changes later, teams can compare downstream use against the previous contract instead of guessing from warehouse query text.
API responses can expose the metric definition, business description, SQL logic, owner, approval context, governance tier, and version context — not just a metric name.
Validation recency, approval status, policy tier, lifecycle state, and version history help an AI agent decide whether a metric is safe to use before acting on it.
Agents can search governed names and definitions, then fetch ownership and lifecycle context instead of choosing from ambiguous warehouse labels.
The contract sits above warehouse and BI implementation choices. Agents, dashboards, notebooks, and internal apps can read the same governed context without changing where computation happens.
Connect your AI agents to governed metric logic. Every agent decision is traceable to an approved definition.
Ensure dashboards and reports reference the same governed definitions. Eliminate the trust deficit.
Query the API from any internal tool. One source of truth for metric meaning across the organization.
Join the companies building a trusted context layer for their AI agents and business teams.
Start Free