All-in-one incident management engineering teams rely on.

Keep your team coordinated from first alert through the retrospective, with AI that actually helps—not just summarizes what happened.

Get started for free Book a demo

Built for the engineers who keep things running.

Respond from where you work.
Declare incidents, assign roles, page responders, update stakeholders, and run the full response—directly from Slack or Teams.
On-call that respects your time.
Schedules, escalation policies, alert routing, rotations, holiday calendars, and live call routing—built to get the right person paged, once.
Learn from every incident.
Retrospectives that generate from the full incident context—no manual reconstruction, no blank page.

Investigate faster with AI that is actually helpful.

AI that correlates alerts, surfaces root cause with confidence scores, drafts status updates, responses to any queries—while your team is in the incident or looking for info about past incidents.

Rootly App Jan 29th at 3:15 PM

Investigating…What I've found so far

Likely root cause

A transient spike in message production is the primary trigger, with the producer rate up to 31 messages at 23:09 UTC, inside the alert's 5-minute window.
Network latency may have contributed, with a spike to 975.63 ms shortly before the alert window.

What happened

The alert triggered at 23:13:58 UTC with kafka.waterdrop.queue.size.p95 at 8.0 messages, above the monitor threshold of 5.0.
Investigation of kafka.waterdrop.produced_async found a production spike to 31 messages at 23:09 UTC, against an average of about 7.
kafka.waterdrop.network.latency.p99 spiked to 975.63 ms at 23:05 UTC, then settled under 50 ms by the time of the alert.
No producer errors kafka.waterdrop.error_occurred were recorded during the window.

Recommendations

Review the threshold on kafka.waterdrop.queue.size.p95, which looks sensitive next to the documented warning level of 250 messages.
Keep watching kafka.waterdrop.network.latency.p99 for spikes between the rootly-workers cluster and the Kafka brokers.

See details and follow along in Rootly AI SRE

Message Rootly

Context your team needs, before they have to ask.

Service catalogs connect your services, teams, runbooks, and dependencies directly to every incident—so responders know what's affected and who owns it, without leaving Slack or Teams.

It connects your entire organization for unmatched context and flexible workflow customization.

Catalogs

Search all catalogs

All catalogs

CatalogEntitiesCreated atCreated byLast edited

CausesSpot trends so you can prioritize areas for post-incident improvement.82025.01.22Tayla Komson2026.02.19

EnvironmentsYour organization's environments.42025.01.23Florance Kalache2026.02.19

FunctionalitiesFunctionalities that can be impacted by incidents.122025.01.26Allan Baker2026.02.19

ServicesThe services that an incident affects or an alert routes to.462025.01.02Allan Baker2026.02.19

TeamsThe teams that an incident affects or an alert routes to.92026.01.12Tayla Komson2026.02.19

TypesThe kinds of incident your teams declare.62026.02.19Florance Kalache2026.02.19

Showing 1 to 6 of 10 results

Your full stack, connected.

100+ integrations including Datadog, GitHub, Jira, Grafana, New Relic, Sentry, Terraform, Linear, and more—Rootly pulls in observability context and pushes updates automatically, so you spend time resolving, not copy-pasting or searching.

- When Incident is Active
- When Type is Security
- When Severity is SEV0
- When Service is Updated
- Send a Slack message
- Fetch GitHub commits
- Pull Datadog metrics
- Create a Linear issue
Less time to resolution.
Every minute of downtime has a cost. Rootly gets the right people paged, context assembled, and response coordinated in the first five minutes — before the incident compounds.
DraftApproved Share linkEdit Send
Email (desktop)Email (mobile)SMSSlack
Some customers are unable to check out. We've identified the issue and are working on a fix. Next update in 30 min.
Sent to 2 audiences
Leadership Customers
Keep everyone informed, automatically.
Rootly handles stakeholder communication during an incident — status page updates, internal notifications, and audience-specific updates generated by AI — so your responders stay focused on the fix, not the inbox.
Rootly App Jan 29th at 3:15 PM
Investigating…What I've found so far
Likely root cause
A transient spike in message production is the primary trigger, at 23:09 UTC within the alert's 5-minute window.
Network latency may have contributed, shortly before the alert window.
Recording in progress
AI that works the whole incident.
Rootly AI joins your incident calls, takes notes, captures decisions, and tracks action items in real time — so nothing gets lost between the call and the retrospective.
MTTR by service
20d 20h13d 21h6d 22h0m
May 19Jun 9Jun 30
Team trends
Over the last 30 days, the average risk level was 44 points.
Reliability data that tells you where to invest.
MTTR, repeat incident rate, alert volume by service, on-call load by engineer — Rootly surfaces the metrics that show where your reliability debt is accumulating, so engineering effort goes to the right place.