Quality AI Release Notes¶
This document provides information on the feature updates and enhancements introduced in Quality AI of AI for Service (XO) v11.x releases.
v11.20.0 December 07, 2025¶
Minor Release
This update includes enhancements and bug fixes. The key enhancements included in this release are summarized below.
Deploy
Enhanced Import/Export for Taxonomy Builder with Saved Filters
Import and Export support for Quality AI deployment enables seamless management and transfer of Quality AI configurations across AI for Service (XO) environments. Users can export the latest taxonomy version with topic hierarchy and resolution settings, then import using Full (overwrite) or Incremental (add-only) modes—maintaining configuration consistency without manual rebuilding. Also, the Advanced Topic Discovery now supports saved slide-out filters for quick reuse of standard filters. Learn more
Configure
QM Express Onboarding Supports AWS Assume Role Access
QM Express now supports IAM Role (Assume Role) authentication for secure, temporary credential access without storing long-term AWS keys. The update includes a new IAM auth type, Role ARN schema, enhanced validation, and S3 service initialization using STS AssumeRole with automatic credential rotation. This strengthens security, enables cross-account access, and aligns with AWS best practices for least-privilege access. Learn more
Analyze
Enhanced Topic, Intent, and Resolution Analytics Across Quality AI
Analytics visibility is enhanced across Conversation Mining, Audit workflows, and Agent Dashboards:
-
Conversation Mining includes optional columns for Configured Intents, Generated Intents, and Overall Resolution, with color-coded topic sentiment and advanced filters for L1–L3 topics, intents, sentiment, emotion, and resolution. Learn more
-
The Audit Screen provides deeper topic-level insights with click-through intents and detailed L3 resolution mapping, helping auditors quickly locate where each topic or resolution appears in the transcript. Learn more
-
The Agent Dashboard includes scrollable Sentiment and Resolution Insights, with averages, trend indicators, and top or bottom-performing topics. App-level controls enable admins to manage agent access to sentiment and resolution data for flexible team visibility. Learn more
API Enhancements
Raw Data API Supports Generated Intents
The enhanced Raw Data API now uses generated intents. Learn more
This document provides information on the feature updates and enhancements introduced in Quality AI of AI for Service (XO) v11.x releases.
v11.19.1 November 19, 2025¶
Patch Release
This update includes enhancements and bug fixes. The key enhancements included in this release are summarized below.
Configure
New AI-Powered Etiquette Evaluation Metrics
New AI-powered evaluation metrics automatically assess agent etiquette across different interaction scenarios, ensuring consistent service quality and providing actionable coaching insights:
-
By Transfer Etiquette Evaluation: It assesses agent conduct during customer transfers, verifying agents inform customers, explain the reason, and confirm understanding before the handoff. Learn more
-
By Hold Etiquette Evaluation: It assesses agent adherence to best practices for managing voice interaction hold scenarios, including notification, appropriate hold time, and clear conversation resumption. Learn more
Enhanced AI Agent Metric
The enhanced AI Agent metric reduces the number of agent calls, supports weight-based breakdowns across multiple evaluation aspects, and improves the design-time configuration experience. It uses a two-level structure (parent metric and weighted sub-metrics with defined criteria/adherence logic) to analyze multiple conversation aspects in one evaluation, providing structured scores and justifications. Learn more
v11.19.0 October 25, 2025¶
Minor Release
This update includes enhancements and bug fixes. The key enhancements included in this release are summarized below.
Analyze
Keyword Search with Audit Filtering
The new Keyword Search Filter feature lets users easily locate and analyze conversations. Users can search for keywords and refine results by speaker, including or excluding keywords. Each search result links directly to the QA audit screen, where keyword matches are highlighted in the transcript, along with speaker labels and QA evaluation details. This ultimately leads to improved audit accuracy, traceability, and filtering flexibility. Learn more
Analytics
Scheduled Reporting for Quality AI
The Quality AI Reporting feature provides comprehensive data export capabilities for AutoQA and Conversation Intelligence insights. This enables QA Managers, Supervisors, and authorized users to generate, schedule, and download detailed reports covering interaction details, agent performance metrics, and adherence to evaluation metrics. Learn more
New Generative AI Features
This release introduces the following two new generative AI features powered by the latest GPT-4o and GPT-4o-mini models with default prompts:
-
By Value Adherence Validation: Check if extracted metric values match predefined reference data and return binary adherence scores.
-
By Value Metric Extraction: Extract metric values from messages by users, AI agents, or humans. Learn more
v11.18.0 September 27, 2025¶
Minor Release
This update includes enhancements and bug fixes. The key enhancements included in this release are summarized below.
Analyze
Topic Discovery
This enables QA, Supervisors, and Business Managers to visualize conversation topics across the contact center, linked to key metrics, such as Average Handling Time (AHT), Sentiment, and Resolution rate. Using custom taxonomies and intuitive bubble charts, users can identify trends, explore topics in-depth, and analyze related metrics and generated intents, making it easier to enhance coaching and drive process improvements. Learn more
Configure
Taxonomy Builder
The builder enables Contact Center Managers, Analysts, and QA teams to create custom taxonomies with three-level hierarchical structures, enhancing conversation categorization. This improves analysis across channels, allowing deeper insights into topic-level metrics and business-specific trends. Learn more
v11.17.1 September 15, 2025¶
Patch Release
This update includes enhancements and bug fixes. The key enhancements included in this release are summarized below.
Analyze
Message-Level Comments in AI-Assisted Manual Audit
Supervisors and QA users can now add comments directly on individual messages, with required name and text fields. The updated comments are displayed both in the Transcript and in the Comments tab. Users can click a comment to navigate to the related message, which enhances feedback visibility and context. Learn more
Conversation Mining Processing Status
In Conversation Mining, the Processing Status now displays a status icon: green (pass) when all processes have succeeded and metrics are being captured; red (fail) when any process fails, and metric tracking is interrupted. Learn more
Conversation Sources
Quality AI Integration with Agent AI
Quality AI can now be integrated with Agent AI on third-party desktops. This integration streamlines the recording process and utilizes metadata through Agent AI's Custom Object field. Learn more