Business metric readiness
Trace governed Revenue and Legal metric dependencies, blockers, and remediation.
Open Data+ → Metric readiness (/app/data-plus/metric-readiness).
The same dependency graph applies to Revenue and Legal governed metrics, dashboards, authored reports, alerts, headless queries and governed AI explanations. It traces metric → dependent metrics → semantic fields → source dataset and permissions → materialized record quality evaluation → remediation. The Legal MVP definitions reuse the existing Legal catalog.
A metric is ready, partial/warning, blocked, or unavailable. An unmapped field, inaccessible required source, empty persisted dataset or failed required validation blocks computation. Stale or insufficient evaluated record coverage produces a warning. Blockers preserve the exact observed failure count, evidence timestamp and mapping/stewardship link when evidence exists. Unknown does not equal zero. A blocked or unavailable metric's value is null, not 0.
The metric gate is reused by cached indicator reads, dashboard/report queries, alert rule evaluation, headless metric queries and governed questions. Dashboards return a non-success state for blocked metric dependencies, and warning metadata for partial ones. AI answers suppress blocked numeric values instead of claiming false certainty.
Prepared dataset snapshots are separate from live metric inputs. A successful recipe does not, by itself, certify the raw source that the current metric uses. Readiness uses actual record-readiness validations against the metric dataset; published recipe versions are preserved as related lineage evidence.
API: GET /api/app/data-plus/metric-readiness, public GET /api/v1/data-plus/metric-readiness, and MCP blocchi.get_metric_readiness. The same tenant/data entitlement and source read policies apply.
Limitations
The graph can explain known registered dependencies, not invent required quality policies that a tenant has not declared. Missing validations remain explicitly untested. Some platform-specific calculated measures have independent execution contracts and may report their own readiness separately. This read does not apply automated record corrections.
See Cleanup Studio and Lineage.
Last updated on