Every Metric Is Auditable
In regulated industries, metrics directly influence alerts, risk scores, and automated decisions. Lao Zi enforces a mandatory approval pipeline — no metric runs in production until a human explicitly approves it.
Mandatory Test Gate
Every DSL expression must be parsed, validated, and executed against real data before it can be submitted for review. No untested logic enters the pipeline.
Human-in-the-Loop Approval
No metric runs in production until a human explicitly approves it. Batch operations let reviewers approve, reject, or deprecate multiple metrics in a single authenticated call.
Complete Audit Trail
Every state transition is recorded with full provenance — who created it, who tested it, who approved it, when, and why. Immutable history for regulatory review.
Full Lifecycle Management
Deprecated metrics can be reactivated. Rejected metrics can be revised and resubmitted. No data is lost — every version and decision is preserved.
Approval Pipeline
Draft
Not in pipelineAuthor creates or edits a metric definition with DSL expression.
Full Audit Trail
Every state transition is recorded with full provenance. In regulated environments — clinical compliance, manufacturing QA, or governance audits — this immutable history satisfies documentation requirements out of the box.
Key Principle
"No untested logic enters the pipeline. No unreviewed changes affect live analysis. Full audit trail for every metric — who created it, who tested it, who approved it, when, and why."
| Field | Description |
|---|---|
createdBy | Who authored the metric |
approvedBy | Who approved it (email) |
approvedAt | When it was approved |
approvalNotes | Reviewer's notes |
rejectionReason | Why it was rejected (required) |
lastTestedAt | When the DSL was last tested |
lastTestResult | Test output: value, severity, SQL, execution time |
deprecatedBy | Who retired the metric |