SS-9172 is delivery health monitoring + a support runbook. It rides on INFI's existing observability — metrics dashboards, logs, and on-call alerts — plus a written runbook for the team. We deliberately don't build a bespoke delivery dashboard; that would duplicate infra INFI already has.
Health metrics & targets (existing dashboards / alerts)
API order-creation success
≥ 98%
Delivery completion
≥ 95%
Quote latency · p95
≤ 500 ms
What it actually is
Metrics tracked against the targets above in the existing dashboards (Grafana / Datadog / SLS) — no new charts to build.
Failure alerts auto-paged to on-call — delivery-creation failures, status-update lag, transient timeouts (cleared on retry).
Per-order lookup for support — the full delivery lifecycle via the existing order tools & logs.
A support runbook for the team: delivery won't start → status stuck → refund questions → escalation path.
Why there's no screen
Metrics & alerting belong in existing observability — reuse the dashboards + on-call routing INFI already runs.
Support already has order lookup; the delivery lifecycle just needs to appear there — no separate console.
The runbook is a wiki doc, not a product screen — and auto-remediation stays out of scope (retries are transient only).
🚦Pilot launch = watch closely, not a new screen: the team monitors the KPIs & alerts through the pilot; transient retries are expected (cleared on retry). Real problems page on-call and follow the runbook.