Conversation Mining - Interactions

The Interactions feature empowers supervisors to review and filter scored interactions in their assigned queues, enabling them to identify crucial conversations for improvement. It offers insights into conversation quality, agent performance, and the customer experience. Supervisors can save custom filters for audits, tailor the view through metadata and columns, apply keyword-based filtering (by speaker, include/exclude words, hide resolved or irrelevant cases), and leverage a timeline-based search with a jump-to-match display for efficient 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.
    

• 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.

• 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.
    

• 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.

• Keywords: Shows detailed keyword-related information to prioritize conversations by relevance to your keyword filters, such as:

  • Total Keyword Match Count: Displays the total number of "Include" keyword matches per interaction. Example: If "refund" appears 1 time, the match count is 1.

  • Match Type Indicator: Shows whether the match is found using “AND” or “OR” logic with color coding.

    • AND: All keywords matched.
    • OR: Partial keyword match.

  • Filter Integration: The applied keyword filters on the Conversation Mining (Interactions) page are saved in sessions. The session preserves filters across pages and views; integrates with other filters using AND logic, and maintains active filter states throughout navigation.
    

• 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.

• Violation: Speech-based violations that occurred.

• Omission: 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

This Filter categories panel allows you to refine and analyze interaction data based on various parameters configured, such as efficiency, experience, behavior, and keywords.

Filter by Efficiency

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

To filter the Efficiency, do the following:

1. Select Channel type (Chat or Voice) to limit interactions to specific modes of communication.
   

2. Select Audit Status to view conversations that are Audited, Assigned, or Not Assigned.
   

3. Enter the Conversation ID, a unique identifier to search for specific conversations (for example, ticket number).
   

4. Enter the Custom Conversation ID by entering a user-defined or custom conversation identifier (for example, external ID mapping).
   

5. From the Queues dropdown, select an interaction by assigned queue or routing group.
   

6. 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.

7. 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.

8. 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

This category captures the emotional and qualitative aspects of the customer journey.

Sentiment Trend

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

• Opening: Select the Starting sentiment trend from the dropdown to define the initial emotional tone of a conversation.

• Closing: Select the Closing sentiment trend from the dropdown to define the final emotional tone of a conversation.

Example: You can choose sentiment values such as Positive, Negative, or Neutral for both the starting and closing points to map the emotional journey across the interaction.

Emotions

Use the Emotions dropdown to select emotional expressions observed during the interaction, whether from the customer or the agent. Examples include Fear, Anger, and Happiness.

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

This category analyzes interaction dynamics and conduct between agents and customers.

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.

Filter by Keyword

This feature allows you to search, filter, and review conversations using specific keywords or phrases. You can filter by included or excluded keywords and by speaker, with results linked directly to the QA audit screen.

Category	Details
Select Conditions (AND/OR)	Choose how multiple keywords in the Include section are combined: AND: All specified keywords must appear in the conversation (conversation-level). Example: - Include: Billing AND frustrated (Both speakers). - Exclude: Resolved (Both speakers). OR: At least one of the specified keywords must appear somewhere in the conversation. Example: - Include: Refund OR “money back” (Customer only). - Exclude: Processed (Agent only), completed (Both speakers).
Conversation Including Keywords	Search conversations for specific keywords or exact phrases, with optional speaker filters. - Keyword Input: Enter a keyword or exact phrase to search. - Speaker Filter: Select who mentioned the keyword: - Agent only: Count only agent mentions. - Customer only: Count only customer mentions. - Both: Count mentions by either speaker (agent or customer). - +Add Keyword: Add up to 8 keywords using the +Add Keyword button.
Conversation Excluding Keywords	Enter keywords or phrases that must not appear in the conversation. You can also select which speaker(s) must not mention the keyword. - Speaker Filter: Choose who must not mention the keyword: - Agent only - Customer only - Both - Add Exclusion: Add up to 5 exclusion keywords using the +Add Exclude Keyword button. Note: - All exclusions use AND NOT logic. If any exclusion keyword appears (according to the selected speaker filter), the conversation is excluded. - Each row includes a Remove button to delete an exclusion keyword.
Audit Navigation & Filtering	- Enables quick filtering and navigation by keyword or QA question. - Supports transcript auto-scroll and match relevance filtering.
Apply Keywords Action	- Click Apply Keywords to activate the filters and display the selected keywords. - Click Apply to run the search and view results based on all active filters.
Filtered Keywords Interactions	Displays the count of filtered interactions (for example, “Filtered Interactions: 22”) containing the selected keywords after filters are applied.

Manage Saved Filters

Apply Filters

• Click Apply to save the filter settings.

• The applied settings, regardless of any category filter selected, are stored as an Unsaved Filter on the Conversation Mining dashboard page.
  

  Note

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

Save Filters

• Click Save & Apply to save the selected filter category details and proceed.
  

• In the Save Filter window, enter the Filter Name, and enable the toggle option Make this the default view if required.
  

• Click Save & Apply in the Save Filter window to set the Unsaved Filter as the default view filter in the Interactions dashboard.
  

View Saved Filters

• Click the newly set default view filter or any Unsaved Filter dropdown to view all the newly created Saved Filters.

• All the Unsaved Filters and their associated Saved Filters are displayed.
  

Clear Filters

• Click Clear to reset all applied filters. This resets all the Filter By options or categories displayed.
  

• Click Clear All to remove or clear all queue filters.
  

Close Filter Panel

• Click the Cross X icon to close the filter panel.
  

View Filtered Interactions

• Click Filtered Interactions to view the real-time conversation interactions count based on your selected or matched filter criteria.

• This count dynamically updates whenever filters are applied or removed.
  

Saved Filters Customization Options

Once filters are saved, the following customization options are available:

• Copy

  Creates a duplicate of the saved filter.

• Mark as default

  Sets the selected filter as the default filter whenever the Call Mining tab is opened.

• Edit Filter

  Allows modification of the details of a saved filter.

• Delete Filter

  Removes the selected saved filter permanently.
  

Send Feedback