Patch Release CX4.10.3.1

Release Name	CX4.10.3.1
Release Ready Date	2025-12-15
Release Summary	Includes improved conversation data retrieval speed and introduced better Instagram message type identification for more accurate routing and analytics. This release also resolves issues with conversation routing after the Routing Engine restarts, ensures reliable processing of team announcements, prevents agent desk refreshes and draft losses caused by large email images, and enhances system stability when handling invalid data.
Upgrade Guide	CX4.10.3 to CX4.10.3.1

New in CX4.10.3

This patch release on CX4.10.3 delivers key performance enhancements and critical bug fixes to strengthen the platform’s reliability, efficiency, and scalability.

Enhancements

Feature(s)	Description
Enhancement in Conversation Manager	To improve query performance, an index has been added on the conversationId field of Activity in the main storage. This enhancement significantly reduces the time required to retrieve conversation data, resulting in faster response times for agents. It also provides better scaling of the system as the volume of conversations grows, supporting business growth without performance degradation.
Segregate Message Types	Adds a flag in the message header to identify whether an Instagram message is a Direct or social media message. This improvement enables more accurate routing, reporting, and analytics. Businesses can now better track the origin of customer interactions, tailor responses accordingly, and gain deeper insights into channel performance.

Resolved Issues

Issue(s)
Queued conversations are not routed after the Routing Engine restarts	Issue: After restarting or redeploying the Routing Engine, both previously queued and new incoming conversations were not routed to agents. Resolution: The restoration logic now sets the mandatory initialization flag for all conversations restored after a Routing Engine restart. This fix ensures uninterrupted service and reliable routing after system maintenance or failures.
Team Announcement	Issue: The Team Announcement Component caused failures during scheduled cron job execution due to a logic error for specific scenarios. As a result, some announcements remained stuck in the Scheduled state and did not transit to Active or Expired as intended. Resolution: The logic now handles all data scenarios correctly. The component now reliably processes announcements during cron jobs, ensuring timely updates to Active or Expired status as expected.
Frequent refresh and draft loss due to repeated socket reconnection	Issue: Previously, agents experienced frequent refreshes during active chats on the Agent Desk or while drafting emails, caused by repeated socket disconnections due to large inline images in emails. Resolution: Inline images are now uploaded to the file engine and replaced with lightweight URLs in the email body, reducing payload size and preventing socket disruptions. This fix improves agent experience by preventing data loss and interruptions, leading to higher efficiency and customer satisfaction.
Object ID is fetched as an Empty string	Issue: Occasionally, empty or invalid strings were fetched from the DB on adding customer attributes, causing pod restarts. Resolution: The component now resiliently handles such cases to prevent system instability.

Issue(s)

Queued conversations are not routed after the Routing Engine restarts

Issue: After restarting or redeploying the Routing Engine, both previously queued and new incoming conversations were not routed to agents.

Resolution: The restoration logic now sets the mandatory initialization flag for all conversations restored after a Routing Engine restart. This fix ensures uninterrupted service and reliable routing after system maintenance or failures.

Team Announcement

Issue: The Team Announcement Component caused failures during scheduled cron job execution due to a logic error for specific scenarios. As a result, some announcements remained stuck in the Scheduled state and did not transit to Active or Expired as intended.

Resolution: The logic now handles all data scenarios correctly. The component now reliably processes announcements during cron jobs, ensuring timely updates to Active or Expired status as expected.

Frequent refresh and draft loss due to repeated socket reconnection

Issue: Previously, agents experienced frequent refreshes during active chats on the Agent Desk or while drafting emails, caused by repeated socket disconnections due to large inline images in emails.

Resolution: Inline images are now uploaded to the file engine and replaced with lightweight URLs in the email body, reducing payload size and preventing socket disruptions. This fix improves agent experience by preventing data loss and interruptions, leading to higher efficiency and customer satisfaction.

Object ID is fetched as an Empty string

Issue: Occasionally, empty or invalid strings were fetched from the DB on adding customer attributes, causing pod restarts.

Resolution: The component now resiliently handles such cases to prevent system instability.