Conversation Mining - Interactions

The Interactions feature lets the Supervisor view and filter scored interactions in assigned queues to identify key conversations for improvement. It provides insights into conversation quality, agent performance, and customer experience. You can save filters for audits and customize the view with metadata and columns to streamline reviews and improve oversight.

Accessing Interactions

Access Interactions by navigating to Quality AI > Analyze > Conversation Mining > Interactions.

Note

• Interactions may take a few seconds to appear after a call ends.

• If any section of the Agent Interactions dashboard elements displays an "NA" status, it indicates that the corresponding data is not yet available on the Audit Allocations page.

• Intents, Topics, Keywords, and Emotions in Conversation Mining are always shown in the assigned default language, even if the conversation is in a different language.

Interactions Dashboard Elements

The Interactions dashboard includes the following elements:

• Agents: Shows the agent's name who last engaged in and ended the interaction. Hover over the agent’s name to view tagged topics and intents.

  • Topic Tags: Shows classified topics for each interaction as tags. Hover over the tags to see all relevant topics discussed in that conversation.
    

  • Intent Tags: Shows the classified intents of each interaction as tags. Hover over the Intent tags to see all relevant intents mentioned in that conversation.
    

• Processing Status: Shows whether the process completion and metric adherence are achieved or not.

  • Pass: Displays a green icon when all processes have succeeded, and metric adherence is captured.

  • Fail: Displays a red icon when all processes have not succeeded, and metric adherence is not captured.
    

• Actions: Allows you to assign the interaction to a bookmark for later reference.
  

  Note

  To view and tag the bookmarks for future reference, you must create the required bookmarks during the Settings configuration.

• Queues: Shows the queue where the system terminated during the interaction.

  Note

  The evaluation form used to score the interaction corresponds to the queue where the system terminated the interaction.

• Kore Evaluation Score: Displays the Auto QA score for an interaction based on the evaluation form completed by agents. Hover over the warning icon to see the agent’s API processing status (Pass, Fail, or Disabled). Click a conversation to open the Conversation Mining page for related Audit, Conversation Details, and Logs.

• Sentiment Trend: Displays the distribution of positive, negative, and neutral tones across a conversation. It tracks sentiment changes during the call, highlights resolution tone with a special scoring method, and summarizes overall sentiment.

To access and review audit interactions:

• Navigate to Audit Allocations to view detailed audit interactions.

• From the Interactions dashboard, click on any agent interaction to open the corresponding AI-Assisted Manual Audit page.

• On the audit page, you can view the full conversation history along with the audio recording.

Interactions Filters

Interaction filters help you to find specific conversations, review agent performance, and identify improvement opportunities. By using the following filters, auditors can dive deeper into conversations and assess agent adherence to quality standards. Filters update automatically based on the default language set.

Columns

You can enable and filter the following fields using the Columns filter:

• Supervisor Auditor Score: Displays the Supervisor Audited score if the interaction has already been audited or evaluated manually.

• Sentiment Score: Displays the system-generated sentiment score for the interaction based on what the customer said.

• Start Time: Displays conversation's start time in a specified format in the Interaction listing page (for example, 24 May 2024, 1:17:10 PM).

• Duration: Displays the call duration (voice and chat), including talk time, hold time, and after-call work time (for example, 0h 6m 25s).

• Moments: Shows the Moments column counts for adherences, violations, and omissions related to the configured metrics of the interaction.

• Emotions: Displays customer or agent emotional states, ranked by interaction duration percentage from highest to lowest. Each customer shows a single emotional pattern for the entire interaction, such as Happy, Escalation, or Confusion.

• Conversation ID: Displays the user-defined identifier for custom tracking.

• Custom Conversation ID: Displays the unique identifier for each conversation, used as the primary key to track and reference specific interactions.
  

When you hover over the listed Moments, the following metrics are displayed:

• Questions Adherences: The By Question Metrics that are met during the conversation.

• Violations: Speech-based violations that occurred.

• Omissions: Metrics not adhered to, including Playbook steps, Dialog tasks, and By Question metrics.
  

Double-clicking on any of the above interactions opens the corresponding AI-Assisted Manual Audit page, where you can view the conversation history and the recording. Learn more.

Bookmarks

Allows you to assign interactions to bookmarks and view all bookmarks assigned to a specific interaction.

Date Range Selection

Provides the option to select the date range for the conversation interactions. The default date range selected is always the last 7 days.

Filters

This provides the filter options to filter the information based on your requirements.

Note

If you attempt to evaluate an interaction not assigned to you, you cannot submit the evaluation.

Add New Filter

The new filter interaction enables you to focus on areas of interest or those with high potential for improvement, where you can save them for audit assignments. It helps you filter out options and identify which interactions went wrong.

Steps to Add New Filter:

1. Click the Filters dropdown shown in the upper-right corner to add a new filter.
   

Steps to Add New Filter:

1. Click the Filters dropdown shown in the upper-right corner to add a new filter.

2. Click the + Add New Filter.
   

3. Select a Filter By category.
   

Filter Categories

The New Filter provides the following three categories of interest.

Filter by Efficiency

This provides an operational view of areas of interest where there is greater potential for improvement.

To filter the Efficiency,

1. Select a conversation interaction Channel type (Chat or Voice).
   

2. Choose the Audit Status if it is Audited, Assigned, or Not Assigned.
   

3. Enter the user‑defined identifier, Conversation ID, for custom tracking.
   

4. From the Queues list, add the Queue names.
   

5. From the Agent Groups list, add the agent group name based on the queue selected.
   

   Note

   You can filter the Agent Groups, which are part of the queues, not based on agents in the agent group that are part of other queues.

6. From the Agents list, add the agent name based on the queue selected.
   

   Note

   You can filter the agents based on the interactions that are part of the queues and the user is part of.

7. Enable either of the following options:

   • Average handling time: Filters interactions based on the start and end of handling time range of interaction.
     

   • Filter by deviation from AHT: Filters interactions by % deviation from the average handling time across all interactions for the respective date range and the interactions that are going wrong.
     

   • Specify the Deviation of the number percentage.
     

   • If No. of Transfers is selected, specify the filters by the number of transfer that occurred within each interaction.
     

Filter by Experience

Avg. Waiting Time

This provides the following filter drop down range selection conditions in seconds:

• None
• Equals
• Greater than
• Greater than equal to
• Less than
  

Sentiment Score

This indicates the positive sentiment score (higher) and negative sentiment score (lower) interactions.

Provides a slider bar to move the minimum and maximum range of interactions.

CSAT

This shows the distribution interactions across the score range that the customer has responded to the feedback service and drilled down accordingly.

Intent

This indicates the underlying cause and customer intent that the conversation pertains to.

Topic

This indicates the subject that a conversation pertains to.

Churn Monitor

This provides the underlying cause and need that a conversation relates to. It indicates the loss of customers over a specific period.

This has the following two options to churn the monitor:

Churn Risk

Provides the extent of customer churn in a given conversation. In this, the Supervisor can view the churn risk % for a given time period.

Note

The customer churn is calculated once per interaction. Customer churn is not to be calculated as a score.

Escalation

This detects the number of escalations raised to the Supervisor by a customer.

Filter by Behaviour

Sentiment Trend

Tracks how the customer's or agent’s sentiment changes throughout the conversation, from start to finish.

• Opening and closing: Select the Opening and Closing sentiment trend from the dropdown list. For example, Positive Trend, Negative Trend, and Neutral.

Emotions

From the Emotions dropdown, select the emotional expressions during the interaction, either from the customer or the agent. For example, Fear, Anger, and Happiness.

Metric Name

This filter enables supervisors to view interactions by specific evaluation metrics. Use radio buttons to filter by Pass or Fail and select a metric from the drop-down list to refine results.

Metric Qualification

The selected evaluation metric appears as a tag below the input field, and you can clear it by clicking X. When opened from the Adherence Heatmap, filters apply automatically. If queues are selected, only metrics from those queues’ forms appear. The filter retrieves interactions where the metric applies, letting you view failed or adhered interactions using radio buttons.

Language

Select the languages to add from the dropdown list.

Empathy Score

This measures the level of understanding and compassion shown by the agent towards the customer situation. Provides the extent of empathy like frustration or displeasure that a customer has shown (negative sentiment). A higher score indicates a more empathetic interaction.

Crutch Word Score

This indicates the extent of filler words (for example, umm, uh, and so on) which is used by the agent. Higher score indicates the higher usage of crutch words.

Agent Playbook Adherence

This indicates the adherence percentage to the Agent AI playbook assigned to that interaction.

Kore Evaluation Score

This indicates the automated QA score associated with an interaction based on the evaluation form assigned to an interactions’s queue.

Manage Saved Filters

1. Click Apply to save the filter settings, which are stored as an Unsaved Filter in the Conversation Mining dashboard.
   

   Note

   If interactions are not used for audit allocation, you can apply the filter without saving. To assign audit allocations, save and name filters for easy reference in future audits.

2. Click Save & Apply to save the new filter with a new name.
   

3. Enable the Save Filter toggle to make the Unsaved Filter for default view in the Dashboard. All the newly created Saved Filters and Unsaved Filters will be tagged under the Saved Filters list.
   

   Note

   The filtered interactions count allows you to verify the interaction count based on the filter selections you make, this count gets dynamically recalculated as and when you update filter selections. By default, the filtered interactions count will be zero until you make the first filter selection.

Saved Filters Customization Options

Once the new filter is saved, you get the following filter options to:

• Copy

  Allows you to create another saved copy of the filter.

• Mark as default

  Allows you to apply the newly created filter as a default filter whenever the call mining tab is opened.

• Edit Filter

  Allows you to edit a saved filter.

• Delete Filter

  Allows you to delete the saved filter.
  

Send Feedback