Version 5.33
July 28, 2026
Enhancements
GenAI Feature Usage Alert Emails and In-App Notifications
Tenant admins now receive automated email and in-app notifications when any GenAI feature usage reaches 80%, 95%, or 100% of its allocated limit, as well as when the monthly token quota is fully consumed. This gives administrators advance visibility into resource consumption so they can proactively manage usage before a feature is suspended for the remainder of the billing cycle.
Reason Selection in Agentic Evaluation
Agentic Evaluation results now include a dedicated reason field for each evaluated question, capturing the specific justification IDs the AI selected in support of its answer. This improves the traceability and auditability of automated evaluation decisions, making it easier for quality analysts to review low-confidence results and validate AI-driven scoring.
Additional Filters in Auto Tagging
Auto Tagging groups can now be scoped with execution conditions based on attached data. Administrators can define one or more attached-data conditions (combined with AND logic, with OR logic across multiple values within the same condition, up to 5 conditions per group) so that a tagging prompt only runs against conversations matching the specified criteria. This prevents unnecessary token consumption on out-of-scope conversations and gives finer control over where automatic tagging applies.
Customized Tagging Period for Auto Tagging
A new "Custom Date Range" option has been added to the Auto Tagging period selector, alongside the existing preset periods. Administrators can now define an exact start date and, optionally, an end date for a tagging group — including open-ended ranges — allowing tighter control over both the token usage and the scope of conversations covered by a tagging scenario.
Conversation Duration Distribution Report
A new "Conversation Duration Ranges" configuration has been added under Settings > Analytics > Definitions, allowing administrators to define between 2 and 7 custom duration ranges (with the last range optionally left open-ended). These ranges now power a new stacked bar chart on the Interactive Reports screen, showing the number of conversations falling into each duration range per department. The chart is fully integrated with existing date and department filters, includes hover tooltips with department, range, and conversation count, and supports drill-down navigation into Interactive Reports.
All AI Insights Added to Ext-API Get Conversation
The GET /ext-api/v2/conversations/{externalId} endpoint now returns the full set of AI Insights for a conversation — topic, summary, conversation result, non-resolution risk, actions, churn risk, positive feedback, negative feedback, and improvement areas — within the conversationSummary object. This allows integrating systems, such as CRM platforms, to surface complete AI-generated insights directly to agents without a separate lookup.
Configurable Acoustic Parameter Analysis for AI Act Compliance
Acoustic Parameter analysis is now fully configurable at the tenant level via a new "Acoustic analysis" toggle under Settings > Analytics > Analytics Features, consistent with the existing consent-required controls for Gender, Emotion, and Sentiment. For newly created tenants this setting defaults to off. When disabled, the analysis service no longer calculates acoustic parameters, related values are hidden in conversation detail analysis results, and acoustic-parameter-based category and evaluation form items are no longer offered for new configurations — while existing categories and evaluation forms that already reference acoustic parameters continue to function.
Data Feeder: New Event Types for Category, Masking, and Evaluation Results
Four new Data Feeder event types have been introduced to give external systems complete, near real-time visibility into conversation-level outcomes without additional API calls:
- Category Match Results (
IConversationIndexedForDataFeeder) — published when a conversation's category indexing completes, containing the conversation ID/external ID and the full list of matched categories with their names, IDs, and matching scores. - Category Definition Updates (
IQueryIndexCompletedForDataFeeder) — published when a category is re-indexed, containing the category ID, name, updated version number, update time, and a descriptive message. - Redaction Results (
MaskingReady) — published when audio and transcript redaction completes, containing the redacted transcript segments (per channel, with timing) and the storage location of the redacted audio file. - Evaluation Results (
EvaluationResultForDataFeeder) — published when a conversation-based or user-based evaluation completes, containing the evaluation ID, evaluated agent, form, total score, and (when applicable) the related conversation.
These events allow external systems consuming Data Feeder to keep category, masking, and evaluation data in sync without polling additional APIs.
Improvements
Category Group Filter on Conversations Screen
A new "Group by" dropdown has been added to the Category filter panel on the Conversations screen, alongside the existing "Sort by" option. Selecting a category group now surfaces that group at the top of the list with a group badge, followed by its categories as a flat list — making it easier to filter and navigate large category sets by group as well as individually.
Increased Character Limit for Department Names
The maximum length for department names has been increased from 50 to 150 characters, both in the Administration > Departments UI and via the External API (POST/PUT /ext-api/v1/departments). This allows longer, more descriptive organizational names to be defined without truncation.
Export Modal UI Improvements
Checkbox labels in the Conversation List export and Evaluation Results export modals have been simplified by removing the repeated "Include" prefix (e.g., "Include user defined categories" → "User defined categories"), with the relevant guidance consolidated into a single description paragraph above each checkbox list. Export functionality is unchanged — only labels and helper text have been updated for clarity.
Minor UI Enhancements
Several minor improvements have been implemented to enhance the user interface. These enhancements make the interface more intuitive and user-friendly, contributing to a better overall user experience.
Minor Bug Fixes
Various minor bugs have been addressed to improve overall system stability and performance. These fixes result in a smoother and more reliable user experience.