Latency budgets
Set a p95 budget per endpoint. Pingcrest alerts on degradation, not just downtime, so you catch the slow bleed before the outage.
budget: p95 < 250ms
window: 5m rolling
breach: 3 consecutive
Built for the on-call engineer
Three things, done properly. Everything else is a webhook.
Incident timelines
Every incident gets an automatic timeline: first failed check, regions affected, status flaps, resolution. Postmortems write themselves.
detected 03:12:41 iad,fra
paged 03:12:49 sev2
resolved 03:26:03 13m 22s
On-call routing
Escalation chains with acknowledgment windows. If the primary sleeps through it, the secondary gets it in 90 seconds.
1. @primary ack 90s
2. @secondary ack 90s
3. #eng-alerts broadcast
incident #4821 · api.acme.dev/health · sev2
03:12:41Check failed from iad, fra. HTTP 503 after 2 retries.
03:12:47Quorum reached. 5 of 12 regions failing. Incident opened.
03:12:49Paged primary on-call. Acknowledged in 41 seconds.
03:18:22Status page updated automatically: degraded performance, API.
03:26:03Recovered. All regions green for 3 consecutive checks. Duration: 13m 22s.
Alerts where your team already lives
First-class integrations, all configurable per monitor and per severity.
>_team chat
>_on-call paging
>_webhooks
>_sms
>_email
>_status pages
>_issue tracker
>_terraform provider
Ship the monitor before the meeting ends
10 monitors free, forever. Solo projects never pay. Teams start at $29 per month.
Create your first probe