What's New — March 16, 2026

This release transforms contact center analysis from a solo activity into a secure, collaborative process. You can now share AI-driven insights from Ender Chat with colleagues, confident that the underlying analytics are more accurate and trustworthy than ever. These updates empower your team to make faster, more reliable decisions based on shared intelligence.

Executive Summary

Ender Chat Collaboration — Securely share your AI-powered chat analyses with specific colleagues or teams, turning individual discovery into team-wide intelligence.

Reporting Integrity Updates — Trust your numbers completely with major accuracy improvements to scorecard calculations and main topic analytics.

WebRTC Agent Health Monitoring — Proactively manage your recording infrastructure with a new dashboard that gives you real-time visibility into the connectivity and health of every agent's WebRTC extension.

On-Demand Chat Uploads — Instantly generate realistic demo conversations with AI or upload your own chat data directly in the UI, accelerating testing and onboarding without needing an API.

Highlights

Ender Chat Collaboration — Securely Share AI-Driven Insights

Previously, your powerful analyses inside Ender Chat were isolated, making it difficult to collaborate on findings. Sharing insights meant resorting to screenshots or copy-pasting, losing the context of the AI-driven conversation.

▸ Collaborate with confidence: Share any chat session with specific users or entire teams. Recipients get a read-only view, so your original analysis remains intact while your findings become team knowledge.

▸ Maintain full control: Your chats are now private by default. As the owner, you decide exactly who sees your work, and you can revoke access at any time. You can also find chats faster by searching for the creator's name.

▸ Secure your workspace: Project folders and conversations are protected by fine-grained access rules, ensuring you only see the chats you've created or that have been explicitly shared with you.

• Before: All analytics users could see all chat conversations, and there was no way to share findings other than by copy-pasting text.
• After: Chats are private by default. Owners can share a read-only version with specific users or teams directly from the chat interface.
• In practice: A QA Manager, Maria, investigates a spike in negative sentiment. She uses Ender Chat to identify the root cause, then shares the chat with a team lead, David, so he has the full context before coaching his agents.

Get started: Open any chat you own, click the Share button in the top bar, and add the users or teams you want to grant read-only access to.

Reporting Integrity — Trust Your Agent & Topic Analytics

Minor discrepancies between detailed session views and aggregated reports could undermine trust in your analytics. Scores in the Score Over Time report didn't always reflect scorecard weights, and sessions could occasionally be assigned multiple "main" topics, skewing filtered results.

▸ Trust your agent scores: The Score Over Time report now perfectly mirrors your scorecard's design. It fully respects point weights and correctly applies critical point rules, zeroing out a session's score when a critical item is failed.

▸ Get precise topic analytics: Every conversation is now guaranteed to have exactly one main topic. When you filter by "Only Main Topic," you can be confident the results are precise and free of unexpected data.

• Before: An agent's score in a report could be inflated (e.g., 89%) because it didn't correctly apply scorecard weights or critical point failures (actual score might be 37%).
• After: Reported scores match the real, weighted calculation you see in an individual review, ensuring consistency across the platform.
• In practice: A director, Sam, reviews the quarterly agent performance report. He can now trust that the agent ranked #1 truly earned their score based on the weighted criteria his team established, allowing him to make confident bonus and promotion decisions.

Get started: This fix is applied automatically. Visit your Score Over Time and Category analytics dashboards to see the more accurate and reliable data.

WebRTC Agent Health Monitoring — Proactively Manage Your Recording Infrastructure

Without visibility into the status of agent browser extensions, troubleshooting missing recordings was a reactive and time-consuming process. It was impossible to know if an agent's extension was misconfigured, offline, or had an expired token until after a problem occurred.

▸ Get at-a-glance visibility: A new health dashboard shows you the real-time connection status of all WebRTC-connected agents, highlighting who is healthy and who has issues.

▸ Troubleshoot faster: Drill down into a detailed list of all agents to filter, sort, and search. Quickly identify agents with stale connections or invalid tokens before they impact data collection.

▸ Reduce alert fatigue: Mute known issues (e.g., an agent on leave) to keep your dashboard focused on actionable problems. The system also automatically clears old alerts once an agent reconnects successfully.

• Before: You had no visibility into agent WebRTC extension health. Issues were only discovered after finding gaps in your call recordings.
• After: A central dashboard in the Marketplace shows a live summary (e.g., "42/45 ok") and provides a detailed, searchable list of every agent's connection status.
• In practice: An admin, Chris, sees the WebRTC app status is "yellow." He clicks to see that three agents have invalid tokens, allowing him to proactively reach out and help them re-authenticate before they even start taking calls.

Get started: Navigate to Settings > Apps > Marketplace > WebRTC. Click the new status chip to open the full observability dashboard.

On-Demand Chat Uploads — Instantly Generate Test & Demo Data

Populating a new or demo environment with realistic conversation data used to require API access or a full production integration. This created a hurdle for validating scorecards, training new users, or preparing for sales presentations.

▸ Generate chats with AI: Instantly create realistic sample conversations for any scenario. Simply describe the topic, choose the language and length, and let AI generate the chat for you.

▸ Upload directly from the UI: You can now upload chat data by pasting JSON or uploading a file directly in the browser, completely bypassing the need for an API key or integration setup.

▸ Iterate quickly: Your draft is preserved after each upload, so you can make small changes and submit multiple conversations without starting from scratch.

• Before: Seeding the platform with chat data required technical effort using the API.
• After: Anyone can add chat data by generating it with AI, uploading a file, or pasting JSON directly into the web UI.
• In practice: A sales engineer, Priya, is preparing a demo for a potential customer in the fintech industry. She uses the AI generator to create ten realistic chats about "disputing a charge" and "reporting a lost card" in seconds, perfectly tailoring the demo environment.

Get started: Navigate to Apps > Upload > Upload Chat to generate or upload your first conversation.

Improvements

▸ Bulk Agent Deactivation — Save time managing your workforce by selecting and deactivating multiple agents at once from the agent group members page. This streamlines the process for shift changes or team reorganizations.

▸ Richer Reviewer Leaderboard — The progress bar in the Reviewers Leaderboard now accurately reflects all completed work by including both reviewed and archived items. Hover over the bar to see a clear breakdown, giving you a truer picture of reviewer productivity.

▸ "Seen" Session Indicator — Sessions you have already opened are now marked with a clear eye icon in the conversation list. This simple visual cue makes it easier to track your progress and avoid re-reviewing work.

▸ Configurable Timezones for WebRTC Recordings — WebRTC recording timestamps now automatically convert from UTC to your organization's configured local timezone. This ensures session times align with your business hours, eliminating mental math.

▸ Improved GoHighLevel Metadata Sync — The integration with GoHighLevel is now more resilient, using batch retrieval to ensure more complete call data and more accurate agent matching, even when individual locations experience temporary issues.

▸ Support for Newest AI Models — You can now use the latest AI models for analysis without interruption. We've resolved an issue that could cause processing errors with certain next-generation model configurations.

Fixes

▸ Improved the reliability of historical data imports from Genesys Cloud, with automatic recovery from temporary API disruptions.

▸ Agent names now display correctly in the conversation list for all call types, including those uploaded via API or WebRTC.

▸ Multi-value settings in Marketplace app configurations are now saved correctly, preventing stale data from persisting after an update.

▸ Enhanced security protections now automatically mask sensitive credentials in settings and encrypt data-at-rest for stronger compliance.