Amazon Connect Updates & Release Notes

You can export cases from the agent workspace to CSV files. Search for cases, select one or more, and choose Export. Use the exported CSV file with spreadsheet applications such as Microsoft Excel or Google Sheets.

When you export cases, you can choose which columns to include, such as status, title, and custom fields. The columns you select determine which case data appears in the file.

You can use the exported CSV file to do the following:

• Generate reports from case data
• Share case data with stakeholders outside the agent workspace
• Analyze case data offline in a spreadsheet application

To use this feature, you must have the Cases - Export permission in your security profile. For more information, see Export cases to CSV.

Connect Customer now enables managers to use generative AI to automatically evaluate the quality of self-service (AI agent) interactions, using the same evaluation framework already available for human agents. As AI agents handle a growing volume of customer contacts, this gives you a scalable way to monitor quality and identify where your AI agents need improvement without manual review of every transcript.

You define evaluation criteria in natural language within evaluation forms, such as “Were all of the customer’s issues resolved by the AI agent?” Connect Customer applies that criteria across self-service interactions automatically, providing detailed reasoning and relevant reference points from the conversation transcript for each score. Managers can review results in aggregate to identify systemic issues, or drill into individual contacts alongside interaction recordings and transcripts.

For more information, see generative AI performance evaluations.

Available in all AWS Regions where Connect Customer performance evaluations are available.

Agents and supervisors now receive automatic notifications when schedules change, rather than having to repeatedly check for updates themselves. You can define rules to send email or text notifications when new schedules are published, when existing schedules are updated, or when an agent’s leave request status changes. This eliminates manual notification effort for schedulers and ensures agents know immediately when something that affects their day has changed.

Available in all AWS Regions where Connect Customer agent scheduling is available.

You can now schedule entire business units, or multiple units that share agents, within a single schedule run. Connect Customer agent scheduling now supports up to 5,000 agents per schedule, up to 350 agents per staffing group, and up to 300 staffing groups per forecast group. Previously, organizations with large or multi-skilled workforces had to split scheduling across multiple runs or maintain separate schedules for shared agent pools, which introduced operational complexity and prevented the optimizer from working across the full workforce.

For more information, see Amazon Connect feature specifications.

Available in all AWS Regions where Connect Customer agent scheduling is available.

When adding non-productive activities such as training sessions or team meetings to agent schedules, Connect Customer now automatically identifies the optimal time for each activity while minimizing impact to service level goals. You specify constraints (anytime within the shift, within a specific time window, or relative to shift start/end) and the system handles placement. Previously, supervisors had to manually scan schedules for available windows, which was time-consuming and often resulted in unnecessary service-level degradation.

For more information, see Add activities to agent schedules.

Available in all AWS Regions where Connect Customer agent scheduling is available.

Connect Customer now provides step-by-step traces showing how AI agents reason, act, and respond during self-service voice interactions, giving you full visibility into every decision an AI agent makes on a customer call. If an AI agent fails to resolve a request, you can drill into the trace directly from the contact details page, alongside the full transcript, and see whether it reasoned incorrectly, called a tool with bad parameters, or timed out waiting for a response.

From the AI agent performance dashboard, you can filter contacts by AI agent name, escalation status, or intent, then navigate into any individual interaction to inspect each reasoning step, including the model used, latency, tool parameter inputs, and model output. No developer access or raw API log parsing is required.

For more information, see AI agent traces.

Available in all AWS Regions where Connect Customer is offered.

Connect Customer now alerts supervisors directly on the real-time analytics dashboard when specific keywords, phrases, or sentiment patterns are detected during live calls and chats. When a customer says something like “cancel my account,” a supervisor sees the alert on the Current agent performance widget, can listen in to the live conversation while viewing the real-time transcript and sentiment analysis, and coach the agent to resolve the issue before the customer churns. You set up alerts by enabling real-time analytics in your flow and creating rules that define which conversational signals should trigger a notification.

For more information, see real-time conversational analytics alerts.

Available in all AWS commercial Regions and the AWS GovCloud (US-West) Region where Connect Customer conversational analytics is offered.

If your contact center handles conversations in multiple languages, agents and managers no longer need to read full transcripts or request translations to understand what happened on a call. Connect Customer now generates AI-powered post-contact summaries in eight additional language families: Portuguese, French, Italian, German, Spanish, Chinese, Japanese, and Korean. Summaries are produced in the language of the original conversation automatically, and also now support non-US English variants including British English and Australian English, reflecting locally appropriate spelling and terminology.

For more information, see post-contact summaries.

Available in all AWS Regions where Connect Customer is offered.

Connect Customer now supports the Interrupt agent flow block

Connect Customer now lets you route a contact to a specific agent even if that agent is at maximum concurrency or in a custom status, ensuring time-sensitive interactions reach the right person without waiting for availability. For example, an agent handling routine service calls can still receive a callback on their personal extension the moment it comes in, then decide whether to place their current caller on hold to take it.

You configure the behavior in a Customer queue flow by specifying the target agent manually or dynamically via contact attributes, and the routing engine handles the override. This works across voice, chat, task, and email channels.

Available in all AWS Regions where Connect Customer is offered. For more information, see Interrupt agent.

Connect Customer now uses generative AI to evaluate self-service interactions

Connect Customer now enables managers to use generative AI to automatically evaluate the quality of self-service (AI agent) interactions, using the same evaluation framework already available for human agents. As AI agents handle a growing volume of customer contacts, this gives you a scalable way to monitor quality and identify where your AI agents need improvement without manual review of every transcript.

You define evaluation criteria in natural language within evaluation forms, such as “Were all of the customer’s issues resolved by the AI agent?” Connect Customer applies that criteria across self-service interactions automatically, providing detailed reasoning and relevant reference points from the conversation transcript for each score. Managers can review results in aggregate to identify systemic issues, or drill into individual contacts alongside interaction recordings and transcripts.

For more information, see generative AI performance evaluations.

Available in all AWS Regions where Connect Customer performance evaluations are available.

Connect Customer now provides schedule update notifications

Agents and supervisors now receive automatic notifications when schedules change, rather than having to repeatedly check for updates themselves. You can define rules to send email or text notifications when new schedules are published, when existing schedules are updated, or when an agent’s leave request status changes. This eliminates manual notification effort for schedulers and ensures agents know immediately when something that affects their day has changed.

Available in all AWS Regions where Connect Customer agent scheduling is available.

Connect Customer now supports up to 5,000 agents per schedule

You can now schedule entire business units, or multiple units that share agents, within a single schedule run. Connect Customer agent scheduling now supports up to 5,000 agents per schedule, up to 350 agents per staffing group, and up to 300 staffing groups per forecast group. Previously, organizations with large or multi-skilled workforces had to split scheduling across multiple runs or maintain separate schedules for shared agent pools, which introduced operational complexity and prevented the optimizer from working across the full workforce.

For more information, see Amazon Connect feature specifications.

Available in all AWS Regions where Connect Customer agent scheduling is available.

Connect Customer now optimizes placement of ad-hoc activities in agent schedules

When adding non-productive activities such as training sessions or team meetings to agent schedules, Connect Customer now automatically identifies the optimal time for each activity while minimizing impact to service level goals. You specify constraints (anytime within the shift, within a specific time window, or relative to shift start/end) and the system handles placement. Previously, supervisors had to manually scan schedules for available windows, which was time-consuming and often resulted in unnecessary service-level degradation.

For more information, see Add activities to agent schedules.

Available in all AWS Regions where Connect Customer agent scheduling is available.

Connect Customer now offers AI agent trace details for self-service voice

Connect Customer now provides step-by-step traces showing how AI agents reason, act, and respond during self-service voice interactions, giving you full visibility into every decision an AI agent makes on a customer call. If an AI agent fails to resolve a request, you can drill into the trace directly from the contact details page, alongside the full transcript, and see whether it reasoned incorrectly, called a tool with bad parameters, or timed out waiting for a response.

From the AI agent performance dashboard, you can filter contacts by AI agent name, escalation status, or intent, then navigate into any individual interaction to inspect each reasoning step, including the model used, latency, tool parameter inputs, and model output. No developer access or raw API log parsing is required.

For more information, see AI agent traces.

Available in all AWS Regions where Connect Customer is offered.

Connect Customer now supports real-time dashboard alerts from conversational analytics

Connect Customer now alerts supervisors directly on the real-time analytics dashboard when specific keywords, phrases, or sentiment patterns are detected during live calls and chats. When a customer says something like “cancel my account,” a supervisor sees the alert on the Current agent performance widget, can listen in to the live conversation while viewing the real-time transcript and sentiment analysis, and coach the agent to resolve the issue before the customer churns. You set up alerts by enabling real-time analytics in your flow and creating rules that define which conversational signals should trigger a notification.

For more information, see real-time conversational analytics alerts.

Available in all AWS commercial Regions and the AWS GovCloud (US-West) Region where Connect Customer conversational analytics is offered.

Connect Customer expands post-contact summaries to eight new languages

If your contact center handles conversations in multiple languages, agents and managers no longer need to read full transcripts or request translations to understand what happened on a call. Connect Customer now generates AI-powered post-contact summaries in eight additional language families: Portuguese, French, Italian, German, Spanish, Chinese, Japanese, and Korean. Summaries are produced in the language of the original conversation automatically, and also now support non-US English variants including British English and Australian English, reflecting locally appropriate spelling and terminology.

For more information, see post-contact summaries.

Available in all AWS Regions where Connect Customer is offered.

Connect Customer now lets you route a contact to a specific agent even if that agent is at maximum concurrency or in a custom status, ensuring time-sensitive interactions reach the right person without waiting for availability. For example, an agent handling routine service calls can still receive a callback on their personal extension the moment it comes in, then decide whether to place their current caller on hold to take it.

You configure the behavior in a Customer queue flow by specifying the target agent manually or dynamically via contact attributes, and the routing engine handles the override. This works across voice, chat, task, and email channels.

Available in all AWS Regions where Connect Customer is offered. For more information, see Interrupt agent.

Connect Customer now brings the same analytics capabilities available for voice and chat to the email channel, giving managers visibility into email interactions without manual review.

Emails are automatically categorized, summarized, and scanned for sensitive data, so you can spot trends, maintain compliance, and evaluate agent performance consistently across every channel.

You can configure automatic redaction of sensitive information like credit card numbers or social security numbers before storage.

You can create rules that act on analytics results, for example auto-categorizing billing disputes or generating follow-up tasks when a complaint is detected.

To enable, add the Set recording, analytics and processing behavior block to your email flows.

For more information, see Set recording, analytics and processing behavior.

This feature is available in the following AWS Regions: US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), and Asia Pacific (Tokyo).

Agents can now update comments, unlink contacts associated with the wrong case, delete cases opened in error, and manage custom related items (such as orders, returns, or invoices) directly from the agent workspace. Previously, correcting these kinds of errors required API calls by an administrator, meaning agents had to submit a request and wait for manual cleanup. Agents can now fix their own mistakes in real time, reducing time to resolution and eliminating unnecessary escalations to internal admin teams.

For more information, see Edit a case in Amazon Connect.

This feature is available in the following AWS Regions: US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), and Africa (Cape Town).

You can now give agents access to their own performance evaluations without also exposing peer evaluations or restricting which contacts they can see. Previously, evaluation access and contact access were coupled, meaning senior agents who needed access to peer contacts for escalation handling or training were forced to either see everyone’s evaluations or see only their own contacts. This launch separates the two, so you can grant broad contact visibility while keeping performance data private to each individual.

For more information, see Evaluation forms security profile permissions.

Available in all AWS Regions where Connect Customer is offered.

You can now restrict the agent login/logout report so team managers only see data for their own direct reports, rather than every agent in the instance. Connect Customer supports both tag-based and hierarchy-based access controls for this report, so you can enforce privacy and compliance requirements that match your organizational boundaries without removing access to the report entirely.

For more information, see Login/Logout reports for agents.

Available in all AWS Regions where Connect Customer is offered, including the AWS GovCloud (US-West) Region.

New Connect Customer Customer Profiles features are now available to you the moment they launch, with no waiting period.

Previously, each new Customer Profiles API required a separate permission update to the Connect Service-Linked Role (SLR) before customers could use it, meaning new capabilities were often inaccessible for weeks after launch even when the underlying feature was ready.

The SLR now uses a consolidated permission model that covers all current and future Customer Profiles APIs automatically.

For more information, see Service-linked roles for Amazon Connect Customer Profiles.

Available in all AWS commercial Regions where Connect Customer Customer Profiles is offered.