Noise reduction
Stage in the incident loop: REDUCEReduce alert noise with rules you can read.
Alert fatigue comes from duplicates, flapping thresholds, cascades and unclear ownership. CallHeim handles each with a published rule.
Example
One noisy source, one incident.
Deduplication and grouping both key off the source. A repeating alert from one tool collapses onto one incident; alerts from different tools open separate incidents unless you route or merge them yourself.
Example alerts for illustration.
The pipeline
By default, seven ordered checks.
Orchestration rules run first, and your own noise rules can add further steps. Suppression runs before everything else, so a suppressed alert cannot open an incident, attach to one or count toward a threshold.
The decision pipeline. By default seven ordered checks run on every alert; orchestration rules run before them and your own noise rules can add more. Most checks divert the alert out of the sequence — a flapping series with nothing open to join is the one exception that still creates an incident.
Maintenance window
Suppress if a window covering this service, team or the whole organization is open right now.
active windowOutcome: suppress
Suppress rules
Suppress if one of your own noise rules matches the alert.
rule matchOutcome: suppress
Fingerprint dedup
Attach a repeat to the open incident in the same service. When the sender supplied a dedup key, that holds at any age. When the identity is only the title or e-mail subject, it holds for 24 hours. A P1 or P2 repeat on an open P3 to P5 incident opens a new incident instead.
open incident; 24 h if title-keyedOutcome: attach
Flap collapse
A series that keeps changing state attaches to the open incident instead of paging again. With nothing open to join, it still opens an incident.
4 changes / 600s (fixed)Outcome: group
Title correlation
Group with an open, still-Triggered alert from the same source and the same service when their titles overlap enough.
similarity ≥ 0.6, 600sOutcome: group
Group rules
Apply your own grouping rules over the window.
600s windowOutcome: group
Create incident
Nothing above matched: open the incident with the source’s severity (or the severity an orchestration rule sets), pick the escalation policy for the service, and page.
elseOutcome: create
Defaults shown above. Orchestration rules run first and your own noise rules can add further steps.
Maintenance windows
Maintenance windows suppress new alerts for a service, a team or the whole organization, and recovery events are exempt.
Schedule a maintenance window with a start and end time.
A resolve arriving during a window still closes the incident it belongs to, so escalation does not keep walking its tiers for a problem that has already ended.
Your own noise rules
Write your own noise rules to suppress, group or hold alerts until a count is reached. Recoveries and acknowledgments from your tools are never suppressed.
Rules run in order and can match on the source, title, body, severity, integration or any field in the alert payload.
Fingerprint deduplication
Each accepted alert gets a stable fingerprint from its organization, its source and its dedup key, falling back to its title. The service is not part of it.
A repeat of the same alert attaches to the incident that is still open for it in the same service, triggered or acknowledged, with no new incident and no new page. When the sender supplies its own dedup key (a PagerDuty dedup_key, an Opsgenie alias, a Datadog alert id, an Alertmanager fingerprint, or a dedupKey that is more than the alert title), this holds however late the repeat arrives. When the identity comes from the e-mail subject or the alert title, the repeat attaches only if the incident was triggered in the last 24 hours; after that it opens a new incident and pages.
A repeat at P1 or P2 is not folded into an open P3, P4 or P5 incident: it opens a new incident and pages. Once an incident is resolved or closed, the next repeat opens a new incident.
An acknowledged incident stays acknowledged when a repeat joins it. A repeat that attaches does not page again, so resolve an incident once its cause is fixed.
Flap detection
A threshold sitting near its boundary crosses it repeatedly. Count the state changes inside a trailing window: at 4 in 600s the series is flapping, and it attaches to the open incident instead of paging again. These thresholds are fixed.
The incident stays open and visible while the series is grouped, so the first page stands and nothing is closed on your behalf. State changes, not events: an alert that simply keeps re-notifying in the same state is persistent, not flapping.
Title correlation
Alerts from the same source and the same service, still open and unacknowledged, group together when their titles reach a similarity of 0.6 within 600 seconds. Title grouping can be switched off for your organization.
Related-incident insights are a separate ranking for people triaging an incident. See how related incidents are ranked →
Every decision is recorded
Each alert the pipeline processes is stored with what happened to it (opened an incident, grouped as a duplicate, or suppressed) and the reason.
CallHeim’s alert handling is rule-based and deterministic: given the same alert, settings and state, it makes the same decision.
Read the rules before you trust them with your pager.
Start a 14-day trial, connect one alert source and send yourself a test page.