What's New — November 16, 2025

This release is about closing the gaps in your quality program. Now you can analyze 100% of conversations—no matter how long—and instantly spot performance trends without ever leaving your dashboard. We've connected your high-level metrics directly to the underlying conversations, enabling you to move from insight to action in a single click.

Executive Summary

Complete AI Coverage for All Conversations — Eliminate blind spots in your analytics. AI summaries and Auto QA scores are now generated for every conversation, including your longest and most complex calls.

At-a-Glance Trend Analysis on Dashboards — Instantly understand if your key metrics are improving or declining. All dashboard widgets now feature trend indicators that compare performance to the previous period.

Interactive CSAT Score Drill-Downs — Move from metrics to action in one click. You can now click any CSAT score on your dashboard to instantly review the specific conversations driving that result.

Automated Training Reminders and Workflow — Keep agent development on track with less manual effort. The system now automatically sends email reminders for upcoming training deadlines, ensuring higher completion rates.

Highlights

AI Coverage for Long Conversations — Eliminate Analytics Blind Spots

Previously, extremely long conversations were sometimes skipped by AI analysis, leaving gaps in your Auto QA and reporting. You had no summary or quality score for your most complex interactions, forcing you to review them manually.

▸ Get 100% AI coverage: All conversations, regardless of length, now receive an AI-generated summary and a complete Auto QA evaluation. The system intelligently preserves the most critical parts of the call—the beginning and end—for analysis.

▸ Eliminate manual reviews for long calls: With summaries and scores available for every session, you can quickly assess long interactions without needing to listen to the entire recording.

▸ Ensure complete and accurate reporting: Your quality and performance reports are now more accurate, with no missing data points due to conversation length.

• Before: Long conversations (over 2,500 words) were silently skipped, generating no AI summary or quality score.
• After: Long conversations are intelligently trimmed and analyzed, so you always get a summary and score.
• In practice: A QA Manager, Jane, running a weekly performance report no longer sees gaps for 45-minute troubleshooting calls and has a complete picture of her team's quality.

Get started: This improvement is enabled automatically. You will now see AI summaries and Auto QA scores for all newly processed conversations in the session detail view.

Dashboard Analytics — Spot Trends and Analyze Performance by the Hour

Understanding performance trends used to require running multiple reports and comparing date ranges manually. It was difficult to see momentum at a glance or analyze performance patterns within a single day.

▸ Instantly spot trends: All dashboard cards and C-Level widgets now display color-coded arrows and percentage changes, showing you how metrics compare to the previous period.

▸ Analyze intraday performance: A new "Hour of the Day" grouping option in charts helps you identify peak hours for volume, duration, or cost, so you can optimize staffing and workflows.

▸ Track new duration and cost metrics: You can now chart Average Duration and Average Cost per conversation, giving you new ways to measure efficiency and find opportunities for improvement.

• Before: Dashboards showed static numbers for the current period only. Identifying trends required manually comparing different date ranges.
• After: Every metric includes a trend indicator showing the change from the prior period, and new metrics like Average Duration can be grouped by the hour.
• In practice: A contact center Director, David, opens his dashboard and immediately sees that Average Handle Time is trending up by 4.2%, prompting him to investigate which topics are taking longer to resolve.

Get started: Trend indicators appear automatically on your dashboards. To use the new chart metrics, simply select Duration, Avg Duration, or Avg Cost from the metric dropdown and choose Hour of the Day in the "group by" options.

Interactive CSAT Drill-Downs — Go from Score to Conversation in One Click

CSAT scoreboards on your dashboard showed you the numbers, but investigating why a score was high or low required you to leave the dashboard and manually search for the right conversations.

▸ Investigate scores instantly: Click any CSAT score or category on a dashboard board to immediately open a filtered list of all the conversations that make up that number.

▸ Pinpoint sources of dissatisfaction: Quickly drill into conversations with low CSAT scores to identify root causes, agent coaching opportunities, or process flaws.

▸ Find best-practice examples: Click on your highest CSAT scores to find and share examples of excellent service with your team.

• Before: CSAT boards were static, displaying numbers without a direct link to the underlying interactions.
• After: All CSAT boards are now interactive, allowing you to click any score to see the matching conversations.
• In practice: Team Lead Sarah notices a drop in the team's exit CSAT. She clicks the "Low Score" segment on her dashboard and instantly gets a list of all relevant calls to review for that afternoon's coaching sessions.

Get started: Navigate to any dashboard with CSAT boards. Click on any score category to be taken directly to a filtered list of the corresponding conversations.

Training Automation — Improve Completion Rates with Less Effort

Keeping agents on track with their training plans often required manual follow-ups from managers to remind them of upcoming deadlines. This administrative work took time away from high-value coaching activities.

▸ Automate deadline reminders: Agents now receive automatic email reminders when their assigned training is approaching its due date, reducing the need for manual follow-up.

▸ Increase completion rates: Proactive, timely notifications help agents stay on schedule and ensure important training isn't missed.

▸ Improve the training experience: Agents can now click to enlarge any image included in a quiz, making it easier to see detailed diagrams or screenshots on any device.

• Before: Managers had to manually track and remind agents about training deadlines. Quiz images were small and hard to read.
• After: The system automatically sends email reminders to agents with a direct link to their training. Agents can click any quiz image to view it full-size.
• In practice: A manager, Tom, assigns a new product training to his team. He can now trust the system to remind agents of the due date, freeing him up to focus on preparing for post-training skill assessments.

Get started: Training reminders are enabled by default and sent automatically. To use enlargeable images, simply add an image to a quiz question; agents will be able to click it during the quiz.

Improvements

▸ Control AI Data Anonymization for More Precise Analysis — You can now choose whether to anonymize conversation data before it's sent to AI models. Disabling anonymization (which is on by default) allows AI to see specific numbers and identifiers, leading to more precise Auto QA scoring and summaries when exact values are critical for evaluation. Find this new toggle in Settings > Auto QA.

▸ Smarter Automation Rule Management — To prevent silent failures, automations (Enders) are now automatically deactivated if a prompt they depend on is archived. The system will also alert you if you try to activate an automation that uses an archived prompt, ensuring your workflows remain healthy and predictable.

▸ Scorecard Partial Scoring Restored — You can once again enable and disable partial scoring options when creating or editing scorecards. The checkboxes for partial scoring are now fully responsive, allowing you to configure more nuanced and flexible scoring criteria for your quality evaluations.

▸ Manage the Internal AI Model from Settings — For greater transparency and control, administrators can now view and configure the AI model used for internal platform operations directly from the Settings > Auto QA page. This setting was previously only accessible by our support team.

Fixes

▸ Richer commenting with emoji support and list formatting, plus more reliable sorting on Prompts and Categories pages.

▸ Report generation and exports to Excel are now more reliable, automatically recovering from timeouts and correctly handling special characters.

▸ AutoQA scoring is more resilient. If an evaluation fails with our faster AI model, the system automatically retries with the more accurate model to ensure a score is always produced.

▸ The reviewer list for disputes now correctly includes all users with permissions to manage disputes, not just those with scoring permissions.

▸ Chart configurations are now preserved even when you rename topics, agents, or customers. Your saved chart selections will no longer break after a name change.

▸ Overall system stability has been improved, with background processing now adapting automatically to server resources to prevent interruptions during high-volume periods.