What's New — March 2026

This month's updates focus on turning your contact center data into trusted, actionable intelligence. We've enhanced core analytics to ensure your reports are built on a foundation of accuracy you can depend on. At the same time, we're making AI-powered research more organized and accessible, helping you move from insight to action faster than ever before.

Executive Summary

Reliable Main Topic Analytics — Ensure your highest-level reports are accurate with a new guarantee that every conversation has exactly one main topic, eliminating data skew from overlapping categories.

Ender Chat Projects — Organize your AI-powered research into structured project folders, making it easy to manage, find, and revisit chat-based analyses for different teams, topics, or review cycles.

Expanded AI Provider Support — Unlock the full power of Ender Chat regardless of your AI provider. You can now use all chat features seamlessly with Azure and OpenAI, not just Google Gemini.

Highlights

Reliable Main Topic Analytics — Trust Your Top-Level Category Reports

Your highest-level analytics depend on accurate categorization. Previously, it was possible for a single conversation to be assigned multiple "main" topics, which could skew reports and make it difficult to get a clean, top-level view of contact drivers.

▸ Get trustworthy data: Analytics filtered by "Only Main Topic" now produce precise and reliable results, ensuring you're making decisions based on accurate data.

▸ Enforce your business rules: Your configured category priorities are now strictly and consistently enforced, guaranteeing the single most important topic is always selected as the main one.

▸ Eliminate confusion: No more unexpected topics appearing in filtered views or dashboards, leading to more confident and straightforward analysis.

• Before: A session could sometimes have several main topics, causing the "Only Main Topic" filter to show more data than expected and diluting report accuracy.
• After: Every session is guaranteed to have exactly one main topic, determined by your category priority rules. Filtered reports are now precise and predictable.
• In practice: A QA Director reviewing the "Top 5 Contact Drivers" report can now trust that each conversation is counted only once under its single, most important topic, giving a true picture of call volume.

Get started: This improvement is applied automatically to all new and updated conversations. No action is required to benefit from more accurate analytics.

Ender Chat Projects — Organize Your AI-Powered Research

As you use Ender Chat for more analyses, your conversation list can become a long, unorganized feed. Finding past research for a specific project or topic required endless scrolling and remembering chat titles.

▸ Group related chats: Organize your AI conversations into named project folders for different teams, review cycles, or research topics to keep your workspace tidy.

▸ Find conversations instantly: Quickly locate specific chats with a title-based search that automatically reveals matches inside collapsed project folders.

▸ Manage your work with ease: Simply drag and drop chats to move them between projects or into a general "Your chats" area for easy organization.

• Before: All Ender Chat conversations appeared in a single, flat list that was difficult to navigate as the number of chats grew.
• After: You can create collapsible project folders to group related chats, with drag-and-drop organization and a powerful search to find what you need.
• In practice: A QA Manager creates a project for "Q1 Agent Coaching" and moves all relevant AI analyses into it, creating an organized hub for their quarterly review work.

Get started: Open Ender Chat. You’ll see new sections for "Projects" and "Your chats." Click "New project" to create a folder and start organizing.

Expanded AI Provider Support — Use Ender Chat with Your Existing Setup

Previously, Ender Chat's powerful analytical capabilities were only available if your organization used Google Gemini as its AI provider. This blocked customers on other platforms from accessing chat-based research and reporting.

▸ Use your preferred provider: Ender Chat now works seamlessly with configurations using OpenAI or Azure, in addition to Google Gemini.

▸ Unlock full functionality: Gain access to all AI-powered chat analysis, research tools, and report generation features that were previously unavailable on some configurations.

▸ No need to switch: You can now enable Ender Chat for your teams without needing to change or reconfigure your organization's underlying AI provider.

• Before: Using Ender Chat with an Azure or OpenAI provider configuration resulted in an error, blocking access to all chat features.
• After: Ender Chat works identically across all supported AI providers (Gemini, Azure, and OpenAI) with no loss of functionality.
• In practice: An admin whose company has standardized on Azure OpenAI can now enable Ender Chat for their QA team, allowing them to ask questions and generate reports without any platform changes.

Get started: No action is required. If your organization is configured to use Azure or OpenAI, Ender Chat will now work automatically.

Improvements

▸ Group Scores Over Time by Agent — The Score Over Time table on the Quality Dashboard now includes an option to "Group by Agent." This allows you to compare individual agent performance trends side-by-side directly in the dashboard, making it easier to spot who needs coaching and who is a top performer.

▸ Clickable Filtered Links in Ender Chat — AI-generated reports in Ender Chat are now more actionable. Reports can include interactive links that take you directly to a pre-filtered list of the exact conversations being referenced, saving you the time and effort of manually recreating filters. Agent mentions also include a quick-access icon to view all of their conversations.

▸ GoHighLevel Integration Enhancements — Your GoHighLevel integration is now more resilient and accurate. Data syncing continues from all valid accounts even if one account's token expires, and you'll see a clear status of which specific accounts need attention. Additionally, agent-to-contact matching is more reliable, using all associated phone number aliases to link conversations correctly.

▸ Seen/Unseen Status in Conversation Lists — It's now easier to track your progress when reviewing conversations. Sessions you've already opened will appear visually muted in the list, while unreviewed conversations remain bolded, helping you quickly spot what still needs your attention without losing your place.

▸ Expanded Analytics Chart Data & Transparency — Analytics charts that use a search filter can now process up to 100,000 conversations (up from 20,000), giving you a more complete view. For queries that exceed this limit, a clear warning icon now appears, ensuring you always know when a chart is based on a partial dataset.

▸ More Accurate Auto QA Scoring — Automated quality scores are now more reliable and consistent. We've improved the underlying logic to produce more accurate scoring results across all supported AI models, ensuring you can trust the outcomes of your Auto QA evaluations.

▸ Proactive WebRTC Extension Health Monitoring — Admins can now proactively identify agent-side recording issues. The platform automatically tracks the health of each agent's WebRTC browser extension, flagging issues like invalid tokens or connectivity problems before they result in missing recordings.

▸ Safeguards for Automated Scorecards — Protect your live automations from accidental changes. The system now prevents you from archiving scorecards or removing scoring points that are actively being used by an Ender, with clear alerts explaining what to do.

▸ Improved API Documentation for Chat Uploads — Integrating your chat and session data is now much faster. The chat and manual session upload API endpoints in our documentation now include detailed descriptions, field-level explanations, and a complete copy-paste request example to accelerate your development work.

▸ Research Time Transparency in Ender Chat — Ender Chat now shows you how long complex AI responses take to generate. A "Researching for..." banner appears on queries that take more than 30 seconds, giving you real-time feedback and clear visibility into processing time.

Fixes

▸ Clearer error messages and automatic formatting cleanup are now in place for user creation and updates.

▸ Analytics filters for Customer Satisfaction (CSI), First Call Resolution (FCR), and Silence duration now correctly handle decimal values without errors.

▸ Ender Chat is more reliable when running complex analytical queries and generating charts from diverse data types.