Analyze effectiveness

Tara uses pre-built analytics dashboards to monitor the automation solution and specific metrics for actionable insights into a Business Process (BP).

View dashboards

To view Tara dashboards, go to Control Tower > Analytics > Dashboards > Payment Screening Alert Review.

note

The Payment Screening Alert Review dashboard calculates metrics for finished transactions only. Thus, not all transactions are considered on the dashboard during the BP execution.

The pre-built out-of-the-box analytics includes the following business metrics:

Chart	Description	Results
False positive (alert rate)	Percentage of messages identified as false alerts	Total volume of false positive alerts
False positive (hit rate)	Percentage of hit alerts identified as false alerts by the ML model and processed automatically	Total volume of false positive hits
Alerts total	Volume of alerts by the ML model decision	Total volume of alerts
Alerts resolved	Volume of resolved alerts by the ML model decision	Total volume of resolved alerts
Alerts generated bar	Volume of alerts by the ML model decision displayed by date	Total volume of alerts
Reason codes	Percentage of resolved alerts based on specific mismatch reasons	Percentage of false positive hits by reason code

You can view the False positive (alert rate) and False positive (hit rate) charts as line charts: False positive (alert rate) line and False positive (hit rate) line, respectively.

View Data Stores

Alert-level analytics Data Store

The pss_messages_analytics_v1 Data Store contains alert-level analytics data.

The available columns are as follows:

• campaignId: unique identifier of the BP installation
• runUuid: unique identifier of the BP execution
• transactionId: unique identifier of a request
• clientId: client executing BP
• messageId: identifier of an incoming alert
• messageType: type of an incoming alert
• alertsCount: number of hits in the alert
• msgDecision: alert decision as a JSON string
• msgReviewDecision: alert gold decision as a JSON string
• reportLink: generated link of a HTML report
• lastModified: timestamp when a record was last modified
• msgFilteredOut: flag if alert was filtered out
• msgDecisionReapplied: flag if a decision on the alert was made by CLDS
• modelId: decision model identifier
• nerModelId: NER model identifier

Hit-level analytics Data Store

The pss_alerts_analytics_v1 Data Store contains alert-level analytics data.

The available columns are as follows:

• campaignUuid: unique identifier of the BP installation
• runUuid: unique identifier of the BP execution
• transactionId: unique identifier of a request
• clientId: client executing BP
• messageId: identifier of an incoming alert
• alertId: identifier of an incoming hit
• entityType: type of a screened entity
• alertDecision: final hit decision
• alertDecisionCode: hit decision code or a name of the rule executed to make a decision
• alertDecisionDescription: hit decision description; currently, the same as the code
• alertDecisionExplanation: hit decision explanation, a verbal detailed explanation of the code
• alertReviewDecision: gold hit decision
• lastModified: timestamp when a record was last modified
• filteredOut: flag if a hit was filtered out
• decisionReapplied: flag if a decision on the hit was made by CLDS
• modelId: decision model identifier
• nerModelId: NER model identifier
• contentFormat: message content format
• tagName: name of the tag where the hit was

Enrichment data analytics Data Store

The pss_de_analytics_v1 Data Store contains enrichment data analytics.

The available columns are as follows:

• type: enrichment data provider
• time: call execution time
• message_id: identifier of an incoming alert
• num_hits: total number of hits in an alert
• num_hits_searched: number of hits that executed the call
• num_hits_decision_changed: number of hits for which the decision was changed because of the call

tip

Also, see View CLDS Data Stores.