We’re building one hub for every monitoring source, alert views your team saves and shares, and on-call. AI will group related alerts and suggest the likely cause.
In development. Pricing will be announced at launch.
team:infrastatus:openseverity:≥warn/var at 96% on db-02Zabbix, 1m×38/tmp at 88% on web-07Alertmanager, 31m×5cache-03Prometheus, 44m×2payments-apiDatadog, 1h×9db-01Grafana, 2h×1ObserveHub will collect alerts from your monitoring tools, group the duplicates, and route what matters to the person on call.
In development. The screens below are examples of what we are building.
One click will acknowledge the alert, stop the escalation, and tell the channel who has it.
Alerts about the same host or service will be grouped before anyone is paged.
Escalation policies will move an alert down the line until someone acknowledges it.
The filters you use every day will get a name, and everyone on the team opens the same list.
Rules will group alerts about the same host. We are building AI that also links alerts from different services when they share a cause, and suggests what started the incident, with the evidence in view.
In development. The screen below is an example of what we are building.
8 alerts, 1 incident, 1 page to Deniz
Waiting for related alerts
Checking 8 alerts…
Log rotation failed on db-02 at 09:12. Old logs filled /var, disk waits rose, and checkout and payments slowed.
Evidence
Suggested next stepFree space in /var/log on db-02, then fix the log rotation job.
Hub first: alerts will be cleaned up before they reach a person, so the engineer on call only sees what needs a human.
This is the design we are building.
No limit on how many you connect.
214 alerts in, 9 incidents out
Rules send each incident to the right team by service, host, tag or severity.
team = infraservice = paymentsseverity ≥ criticalSchedules, rotations and multi-level escalation decide who is notified, and how.
Push, email, SMS, voice call, Slack and Microsoft Teams. Acknowledge from any of them.
We are building schedules with weekly, daily and follow-the-sun rotations, and overrides for when someone swaps a shift. The whole team sees the same calendar.
In development. Example schedule, starting from today.
DDeniz is on call now. Arda is next.
Example weekly rotation: Deniz is on call this week and Arda is next.
Normal alerts reach the person on call by push and email. SMS and voice calls are kept for critical alerts, so a ringing phone means something.
Planned.
No SMS, no call. It waits in the app until someone picks it up.
Acknowledged at 4:12, calls stop.
Connect the monitoring tools you run. No plan limits on how many.
Slack and Microsoft Teams are planned as notification channels. Anything that can send a webhook or an email will be able to connect, including New Relic, Dynatrace, Elastic, Sentry, CloudWatch, Azure Monitor, Checkmk and Uptime Kuma. Logos belong to their owners and mark integrations we are building.
Opsgenie shuts down on . If you want to stay with a focused alerting and on-call tool, this is what we are building for the move.
Importer in development. Tell us what you use today when you join early access.
Flat and transparent. You should be able to read the price of an on-call tool in one line.
No. ObserveHub will be a hosted (SaaS) service only.
In the EU.
We will invite the waitlist in groups as features are ready. Join early access and tell us what you use today; it helps us decide what to finish first.
Tell us a little about your team. We will invite the waitlist in groups and use your answers to decide what ships first.