Anomaly Detection

AutoRABIT Vault Anomaly Detection User Guide

Purpose

AutoRABIT Vault Anomaly Detection monitors configured Salesforce data objects and metadata types for unusual changes. The feature helps identify unexpected activity, review affected records, compare snapshots, and roll back selected data changes where required. The workflow begins with configuration, continues through dashboard monitoring and anomaly review, and ends with rollback or comparison result tracking.

Workflow Covered

• Configure anomaly detection for data and metadata.
• Select notification and exclusion settings.
• Monitor active or paused detection status from the dashboard.
• Review anomaly details and compare detected records.
• Submit rollback jobs and review rollback outcomes.
• Use job history, compare labels, export, and field-selection controls.
• Pause, stop, or restart anomaly detection.

Anomaly Detection is configured from the Anomaly Detection workspace. The configuration defines the source org, monitored data objects, monitored metadata types, threshold percentages, notification recipients, and excluded change owners. Once the configuration is saved, AutoRABIT Vault begins evaluating the selected scope based on the scheduled detection cycle.

Anomaly Detection landing page before configuration

Config Creation window with Data threshold settings

Data object selection for anomaly monitoring

Config Creation window with Metadata threshold settings

Metadata type selection for anomaly monitoring

Email Notifications and External Changes of Users options

User Details selection for excluding internal application users

External Changes of Users option after selecting AutoRABIT Vault users

Salesforce User Details selection for excluding Salesforce users

Monitoring the Dashboard

After configuration is saved, the dashboard displays the detection status, source org, date range, and anomaly summary cards for Data and Metadata. The status indicator shows whether anomaly detection is active or paused. The dashboard presents detected data changes by severity so the investigation can continue from a summarized view into detailed records.

Anomaly Detection dashboard after configuration is saved

Active anomaly detection status on the dashboard

Data anomaly summary chart for the selected object

Data anomaly summary chart with severity filter

Reviewing Data Anomaly Details

The Data anomaly details view lists detected records for the selected source org, object, and detection period. Records can be reviewed in the table, selected for rollback, or compared for deeper field-level analysis. The system presents available actions based on the selected records and the current result state.

Anomaly Details page with detected records

Record selection for rollback action

Rollback prerequisite and behavior information

Rollback is initiated from the anomaly details or result views after eligible records are selected. AutoRABIT Vault presents a rollback summary before submission, including rollback options and the selected record count. Once confirmed, a rollback job is created and tracked from the Anomaly Rollback section until completion.

Rollback Summary window with selected records and rollback preferences

Rollback job submitted confirmation

Anomaly Rollback job list

Rollback job actions and progress tracking

Rollback result details with Follow Records tab

Rollback result details with Outcome Records tab

Completed rollback status in the rollback job list

Comparison results are accessed from Anomaly Detection Job History. The results view shows detected changes by compare label and object. Field-level changes are highlighted in the results table, and detailed record comparison is available through the View Record action. Export and field-selection options support focused review of the result set.

Return to the Anomaly Detection dashboard after rollback review

Anomaly Details page with records selected for comparison

Comparison job submitted confirmation

Anomaly Detection Job History list

Job history action for viewing comparison results

Anomaly Detection Results page with comparison records

View Record action in the comparison results table

View Record window showing field-level snapshot comparison

Export option in Anomaly Detection Results

Export window with available export scope options

Choose Fields option in Anomaly Detection Results

Fields selection window for result display

Records selected for rollback from Anomaly Detection Results

Rollback action from Anomaly Detection Results

Rollback fields selection window

Rollback job submitted from comparison results

Anomaly Detection Job History maintains the comparison and rollback activity initiated from the anomaly workflow. Each job entry provides status, timing, and action controls. Compare label actions open detailed field-level status information for the selected compare run.

Job History list after rollback submission from results

Job History action for compare label review

Compare Label window showing field-level status

Job History action for detailed compare label review

Compare Label window with scrollable field status list

Pausing, Stopping, and Restarting Detection

Anomaly detection can be temporarily paused until a selected date or permanently turned off. A confirmation message appears before permanent changes are applied. When detection is restarted, the dashboard returns to an active monitoring state and allows the detection date range to be selected again.

Date picker for changing the dashboard date range

Pause anomaly detection date selection on the dashboard

Save action for pausing anomaly detection until a selected date

Active status after date-based anomaly detection control

Turn off anomaly detection permanently option

Confirmation window for turning off anomaly detection permanently

Stopped anomaly detection status after permanent turn off

Confirmation window for restarting anomaly detection

Date picker available after anomaly detection is restarted

Result

After the workflow is completed, AutoRABIT Vault maintains the anomaly configuration, displays the current monitoring state on the dashboard, stores comparison jobs in job history, and tracks rollback jobs separately. This provides a controlled path to identify suspicious changes, verify field-level differences, and restore selected data where required.