CX Release Notes

CX4.7

Maintenance releases of CX4.7 will be available until 24 Jan 2028

QA PASSED

24 Jan 2025

• Security Enhancements

• Deployment Optimization

• Performance Improvements

• Agent Experience Improvements

• Supervisor Experience Improvements

• ETL Enhancements

Enables two-factor authentication for CX users (Agent, Supervisor, or Admin) to ensure authenticity more securely at the login screen on CX user apps (Agent Desk, Unified Admin)

This feature enables admins to mark customer attributes as PII. All attributes marked as PII are restricted and displayed only to the users who can view the attribute's contents.

Agent Teams in EFCX are set up in unified admin. See Team Configuration in Unified Admin  

Features

Description

Helm

Helm is a templating engine for Kubernetes to minimize the effort of customizing and updating several kubernetes manifests using helm charts. Additionally, the helm charts provide templating which make it easy to interpolate deployment variables, upgrade, downgrade release using simpler and convenient steps.

In previous releases, Agent and Customer SLA notifications (start, stop, threshold) were being published on Conversation Topic which was impacting the customer and agent messages delivery throughput.

Separate Queues have now been introduced for Agent Manager and CCM to publish and consume SLA notifications.

MongoDB replica-set connectivity support has been introduced with read preference, where write operations will be performed on the primary instance, and read operations will be redirected to the secondary instance, thus avoiding the overload on the primary instance.

The application was experiencing gradual memory consumption over time due to improper memory management and occupied memory was not being released. Memory Leak was identified and fixed in core components which resulted in stable memory usage under load.

High CPU Spikes were also observed in previous releases under peak operations. Threadpool management has been improved to reduce CPU Spikes under high load.

ActiveMQ Consumers will auto-scale now based on the system load thus improving the overall message throughput.
Currently only auto-upscaling is supported. Auto-downscaling doesn’t happen at the moment (the memory footprint of this is very low, so it has no impact on performance as of now) and will be addressed in future releases.

Displays call hold time on the CTI toolbar when the call is on hold, and the hold timer starts from 0 on every hold.

Allows agents and supervisors to edit their messages within the Agent Desk.

Enables agents to have a smooth call experience by managing microphone access and preventing unclear call drops by clearly indicating the need for mic permission.

Provides an improved user experience of Dialpad-Full View

Handles issues that occur before a call is accepted because of browser refresh, browser closure, and RONA.

A simplified user experience where agents can easily transfer conversations to a named agent, without bothering which queue the agent is a part of.

Features

Description

Barge-in for CX Voice

In addition to barge-in chat conversations, supervisors can now monitor and barge in ongoing voice calls.

Introduces a custom Python-based ETL (Extract, Transform, Load) tool enabling users to extract data from various sources, perform transformations, and load the results into different targets. This is designed for easy deployment, customization, and orchestration, making it ideal for scalable data workflows.