USE CASE · OPERATIONS

The alert goes to the right person, not into the noise.

Your monitoring stack fires a webhook; SendNomi routes the alert to the right team by severity and label, distributes digests on schedule, and runs incident response on the record from start to finish. Operations email is either noise or a lifesaver — the difference is the routing rule.

P1–P3
severity + label based team routing
12 months
timestamped alert log — who, when, why
5 steps
incident response chain — detection to review
01 / Route

Not noise — one email to the right team.

A Prometheus threshold breach or a Grafana alarm arrives as a webhook; the rule looks at severity (P1–P3) and labels to pick the target team. The on-call engineer gets a formatted email with a context summary, and every alert lands in a timestamped record. SendNomi sits in front of your existing monitoring stack — no SDK required.

ALERT ROUTING
  • Event A threshold breach or alarm arrives as a webhook.
    POST /webhooks/alerts
  • Rule Severity + label matching picks the target team.
    severity=P1 → team: on-call
  • Send A formatted alert email with a context summary.
    template: alert-context
  • Record Who, when, why — a timestamped log.
    retention: 12 months
Prometheus · Grafana · Alertmanager · generic webhook
02 / Digest

The noise settles into one morning report.

Daily and weekly digests generate themselves on a cron trigger: a template-based report goes to the team as a single email, with no manual collation. A new route, a maintenance window or a release note travels the same channel — published in one click to a role-based audience, so the whole team hears it at the same time.

SCHEDULED REPORTING
  • Daily log digest A template-based report reaches the team every morning.
    trigger: cron · daily
  • Weekly digest The week’s metrics merge into a single email.
    trigger: cron · weekly
  • System announcement Deploy, maintenance, route — published in one click.
    target: role-based
No manual collation — the report renders from a template
03 / Manage

An incident is not only resolved; it is announced, tracked, reviewed.

The moment a P1 alarm opens, the on-call engineer has the context summary; stakeholders get status from a single source, and progress notes go out on their own at set intervals while resolution runs. When the incident closes, the impact window and summary are distributed; the post-incident review timeline fills itself from the audit log. No step falls off the record.

Incident response chain — 5 steps
  1. DETECT Incident opened

    P1 alarm; the on-call engineer gets an instant email with the context summary.

  2. ANNOUNCE Stakeholder notice

    Affected teams and managers get status from a single source.

  3. UPDATE Progress note

    Automatic status updates at set intervals while resolution runs.

  4. CLOSE Resolution note

    The impact window and a short summary go to all stakeholders.

  5. REVIEW Post-incident report

    The timeline fills itself from the audit log.

EXAMPLE SCENARIO

The first seconds of a P1 alert.

How an alert enters as a webhook and becomes a traceable record in a representative setup — latency values are for a typical configuration and vary with rule complexity.

EVENT 0 sec the webhook lands
DECISION < 100 ms severity + team match
DELIVERY < 3 sec signed email to on-call
RECORD 12 months timestamped retention
Sources: Prometheus · Grafana · Alertmanager · generic webhookNo SDK required

The alert email carries a context summary: which service, which threshold, since when.

GET STARTED

Your first send within minutes.

Create your free account — 500 sends per month, forever.