AIOps Value dashboard

Related topics

As an operator or Site Reliability Engineer (SRE), use the AIOps Value dashboard to view the key performance and operational metrics, and the top offenders for events, and assess how effectively situations help improve incident handling, reduce noise, and accelerate issue identification and resolution over time. Top offenders are events that cannot be effectively correlated into situations due to a missing node or service context.

Operators or SREs can use this dashboard to perform the following actions:

Get real-time visibility into how BMC Helix AIOps improves service reliability and operational performance.
Track efficiency trends in event correlation, detection, and resolution over time.
Make data-driven decisions to proactively optimize service performance and operational efficiency.
Use an out-of-the-box, ready-to-use dashboard without requiring any custom setup.

Scenario

Jim is an SRE at Apex Global and wants to analyze BMC Helix AIOps performance indicators for the last 15 days to understand which issues and services had the greatest impact and how effectively situations reduced operational noise.

He logs on to the dashboard and navigates to the AIOps value dashboard. He specifies 15 days as the date range to view and analyse data.

He observes that event noise is reduced by 70.9% and incident noise by 78.4%, indicating that event correlation is significantly lowering alert volume and preventing unnecessary incident creation. The Mean Time to Identify (2.07 minutes) shows that issues are detected quickly, while the Mean Time to Resolve (13.6 minutes) reflects relatively fast remediation.

To further reduce MTTR, Jim reviews the Operational KPIs and sees that although most events have node (91.5%) and service (88%) associations, some gaps remain. The Top Offenders section highlights specific event classes and source hosts generating events without node or service context.

Using these insights, Jim identifies data and modeling gaps to address, thereby improving correlation accuracy, reducing standalone events, and further accelerating issue resolution.

The following image shows the AIOps Value dashboard with sample data:

To view the AIOps Value dashboard

Log in to BMC Helix DashboardsBMC Helix Dashboards.
From the navigation menu , click Dashboards.
In the Service Monitoring folder, click AIOps Value Dashboard.

Tip: Quick access from the home page

To quickly open the dashboard from the home page, mark it as a favorite by using the star icon. Additionally, after you open a dashboard, it is available under Recently viewed dashboards on the home page.

Panels in the AIOps Value dashboard

The following table describes the panels in the AIOps Value dashboard:

Panel	Description	Example
Dashboard filter	By default, the dashboard displays the data for the last 30 days. You can filter the data by using the time range global filter for up to the last 90 days.
Service Name filter	Allows you to select a specific service, multiple services, or all services. Except for the panels in the Top Offenders section, data in all panels on the dashboard is filtered based on the selected services. By default, the dashboard considers all events for calculations, except the following classes and severities: Classes: INCIDENT_INFO AUTOMATION_STATUS_EV Situation Change Prediction Severity: INFO
Business KPIs
% Event Noise Reduction	Displays how the overall event noise is reduced by correlating related events into situations. Formula: Event noise reduction (%) = (E - (S+SA)/E) × 100 Event noise reduction (%) = (Total number of events - (Number of situations formed+Number of standalone events)/Total number of events) × 100 where: E = Total number of events S = Number of situations formed SA = Number of standalone events (that are not part of any situation)
% Incident Noise Reduction	Displays how the number of incidents created from correlated events is reduced. A higher incident noise reduction percentage indicates that duplicate or unnecessary incident creation is reduced by grouping related events into situations. This percentage helps teams focus on fewer, more meaningful incidents, improving productivity and service response time. Formula: Incident Noise Reduction (%) = ((Events qualified for incidents−incidents formed)/Events qualified for incidents)×100 where, Events qualified for incidents = Number of events that meet the criteria to trigger incidents Incidents formed = Number of actual incidents created after correlation If the policy for incident criteria is disabled, the incident noise reduction value is zero.
Mean Time To Resolve	Displays the average time to resolve an issue after it is identified. It reflects the end-to-end efficiency of the operations team in remediating the issue and restoring normal service. In BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps, MTTR is measured from the moment a situation is identified (or a ticket is created) until the situation is closed or the fix is applied and confirmed. Lower MTTR values indicate faster remediation and improved operational performance. Important: MTTR is a global business KPI computed solely from situations. Even when filtering by specific event types (for example, Anomaly), the MTTR value displayed represents the global situation-level metric across all event types. If no situations are formed for the selected event type, the displayed MTTR still reflects the overall situation performance, not event-specific data. Formula: MTTR = ∑(Time resolved−Time identified)/Number of situations where, Time identified = Timestamp when the correlated situation is formed Time resolved = Timestamp when the situation is closed Note: If the selected Event Type filter shows “Events part of Situation = 0”, it means no situations were formed from that event type. However, because MTTR is a global situation-level KPI, the value displayed still reflects overall situation performance and not metrics for the selected event type.
Mean Time To Identify	Shows the average time taken to detect and correctly identify an issue or anomaly in the monitored environment. In BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps, this metric is calculated from the moment the first relevant event is ingested until the events are correlated into a situation and classified as actionable. A lower MTTI indicates faster detection, better operational efficiency, and effective event correlation in situations. Important: MTTI is a global business KPI computed from Situations only. Even when filtering by specific event types (for example, Anomaly), the MTTI value displayed represents the global situation-level metric across all event types. If no situations are formed for the selected event type, the displayed MTTI still reflects the overall situation performance, not event-specific data. Formula: MTTI = ∑(Time identified−Time first event occurred)/Number of situations where, Time first event occurred = Timestamp when the first relevant event for that situation was ingested. Time identified = Timestamp when the correlated situation is formed. Note: If the selected Event Type filter shows "Events part of Situation = 0", this indicates no situations were formed from that event type, but the MTTI value displayed is still the global situation metric, not specific to that event type.
Operational KPIs
Total Events	The total number of events ingested by BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps within the selected time range. This time range includes all events, whether correlated into situations or standalone.
% Events with Node Association	The percentage of events that are linked to a node (infrastructure element) in BMC Helix Discovery and a service model. These events are mapped to infrastructure components.
% Events with Service Association	The percentage of total events that are associated with a service. %Events with Service Association=(Events with Service Association/Total Events)×100 A higher percentage indicates better service-context awareness in event data.
Total Situations	The total number of situations created by BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps by correlating related events using ML and topology-based analysis.
% Events part of Situations (Event Correlation Coverage)	The percentage of total events that are correlated into situations. %Events part of situations=(Events part of situations/Total events)×100 A higher percentage indicates better event correlation coverage and lower operational noise.
Top Offenders
Events without Node Id	Displays events that are not associated with any node. Events are grouped by event class and shown as a pie chart, where each slice represents the percentage contribution of a class to the total events without a node association.
All class events without Node Id	Displays events of the selected class from the Events without Node ID panel that are not associated with any node. These events are further grouped by the source_hostname topology facet and shown as a pie chart. Each slice represents the percentage contribution of a source host to the total events without a node association for the selected class.
Events with Node Id but without service	Displays events that are associated with a node but not linked to any service. Events are grouped by event class and shown as a pie chart. Each slice represents the percentage contribution of classes that are not qualified for service-level correlation.
All class events with Node Id but without service	Displays events of the selected class from the Events with Node ID, but without the Service panel, that are associated with a node but not linked to any service. These events are further grouped by the source_hostname topology facet and shown as a pie chart, where each slice represents the percentage contribution of a source host to the total events with a node association but without a service association for the selected class.
Raw data	Displays total events with the following details: Class: Represents the event classification used by BMC Helix AIOps to identify the type of event (for example, alarm, change, prediction, or incident-related information). Source Hostname: The name of the host or source system from which the event originated. This value is used to identify the originating source and group related events. Node Id: The unique identifier of the node (host, VM, or device) from which the event originated. Service Key: The internal key used by BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps to link the event to a specific service entity in the database or service model. Service Name: Name associated with the service created in BMC Helix AIOps or discovered in BMC Helix Discovery. Impacted Service Name: The name of the business service impacted by the event. Impacted Service Key: The name of the business service determined to be impacted by the event based on BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps correlation and dependency analysis. Creation time: The date and time stamp of the event creation. Arrival time: The date and time when the event was received by BMC Helix AIOpsBMC Helix AIOpsBMC Helix AIOps. Incident Id: The identifier of the incident created from the event or Situation, if applicable. This field is populated when the event results in an incident. Message: The descriptive message or text associated with the event, providing additional context about the issue or anomaly. Severity: The severity level assigned to the event (for example, Critical, Major, Minor, or Info), indicating the urgency and impact of the issue. Errors: Indicates whether any errors were encountered in the event payload, such as missing or malformed fields, incorrect associations, or ingestion issues. Identifier: A unique ID assigned to each event. Situation: Indicates the Situation to which the event is correlated. If populated, it confirms that the event is part of a correlated Situation; if empty, the event is treated as a standalone event. Rstat: Indicates the relationship status or state of the event in relation to other events or situations. Tenant ID: Identifies the tenant or account to which the event belongs. Tenant Name: The display name of the tenant or account associated with the event.

To improve the noise reduction accuracy in dashboard calculations

From the AIOps Value dashboard, click the Dashboard Settings icon.
Select Variables.
Click the eventType variable.
In the Custom all value field, add the event classes that you do not use for forming situations and want to exclude from the dashboard calculations.
Save the changes.

Important

When excluding event classes from the dashboard calculations, specify them using the following format in the Custom all value field:

-class:"INCIDENT_INFO" AND -class:Situation AND -class:Change AND -class:Prediction AND -severity:INFO

If a class name includes special characters, make sure it is enclosed in double quotes so that the filter is parsed correctly.

Excluding irrelevant event classes ensures that the noise-reduction metrics more accurately reflect the effectiveness of situations, rather than being influenced by events that are intentionally not correlated.

FAQs

Why do I see MTTR or MTTI values when filtering by Anomaly event type, even though no situations are formed (Events part of situation=0)?

MTTR and MTTI are global business KPIs calculated from all situations, regardless of the Event Type filter selection. These metrics are situation-level performance indicators and are not filtered by event type.

When you filter by an event type that does not form situations (for example, INFO-severity Anomalies), the operational KPIs will correctly show Events part of situation=0 for that event type. However, the business KPIs (MTTR or MTTI) continue to display the global situation metrics across all event types.

This is because:

Business KPIs measure overall system performance and situation handling efficiency.
Operational KPIs provide event-type-specific metrics.
Dashboard panels cannot be conditionally hidden based on filter selections.

Which metrics are affected by the Event Type filter?

The Event Type filter affects operational KPIs like Total Events, Events part of situations, Events not part of situations, Events with Node Association, Events with Service Association, Events with Errors, and the raw data table.