This feature will be available in Public Preview beginning March 16, 2026.

Empowering contact centers to safely handle sensitive, high‑trust customer journeys 

Delivering secure, compliant, and seamless customer experiences is no longer optional — it’s foundational. Across industries such as financial services, healthcare, and public sector, organizations must enable customers to complete high‑trust actions (like payments or identity verification) without exposing sensitive information to agents or core contact center systems. Secure contact center customer journeys enabled through capabilities like Secure Consult & Transfer bring this capability to life by allowing agents to involve external secure endpoints in the conversation while maintaining strict privacy boundaries. 

Today, we’re excited to highlight how Secure Consult & Transfer modernizes sensitive interactions and prepares organizations for the next generation of compliant service workflows. 

How It Works 

Secure numbers are created by applying specific settings to a contact. During runtime, the platform enforces the necessary protections instantly — without requiring manual intervention. The settings are flexible, and can be applied to only consult or transfer, or both. For consult and transfer the administrator can decide to follow workstream recording & transcription settings, stop recording but continue transcription, or stop both transcription & recording. Additionally, during consult, administrators can choose to either put their representatives on hold, or follow workstream settings to have the customer placed on hold with the representative able to take them off. 

Key Benefits at a Glance 

• Protect sensitive customer data without disrupting workflows. 

• Enable secure payment and verification scenarios using external trusted endpoints. 

• Automate compliance controls — recording and transcription management happens instantly and safely. 

Secure Consult & Transfer ushers in a new era of secure, compliant customer interactions — one where sensitive workflows can happen inside the call experience without adding risk, friction, or operational burden. 

If your organization handles sensitive customer actions, now is the time to explore how this capability can strengthen trust, reduce risk, and streamline your service operations. 

Accelerating Secure Contact Center Customer Journeys with DTMF Broadcast 

A modern foundation for high‑trust, compliant, real‑time voice interactions 

As organizations modernize customer engagement, the need for secure and friction‑free voice workflows has become essential—especially for processes that rely on keypad inputs, such as payment authentication, IVR navigation, or identity verification. Traditional DTMF forwarding approaches rely on slow relays, creating latency, reliability gaps, and compliance concerns. 

DTMF Broadcast introduces a new, faster way for participants in a call to share DTMF tones in real time— addressing these gaps by sending DTMF tones from one participant instantly to all non‑hold participants, far faster than traditional forwarding. 

This includes all legs of the call, meaning that representatives can more reliably send tones to external endpoints, and if representatives drop off after a transfer to an external endpoint, customers will still be able to send DTMF to that endpoint, navigating IVRs independently. 

How to Enable DTMF Broadcast 

The toggle to enable DTMF Broadcast for an organization is in the Copilot Service Admin Center under Support Experience->Workspaces->Voice Experiences. 

The post Reimagining Secure Customer Interactions with Secure Consult & Transfer  appeared first on Microsoft Dynamics 365 Blog.

With the latest July 2025 updates to the Customer Intent Agent, Microsoft is accelerating the journey toward fully autonomous contact centers. These enhancements empower organizations to leverage AI agents that not only understand customer intent but also automate the entire resolution process. This increases the availability of human service reps to focus on high-value interactions. 

The July preview introduces powerful new capabilities for the Customer Intent Agent, building on the foundation of intent mining, dynamic interviews, and knowledge-driven automation. Now, organizations can achieve more accurate intent detection, seamless handoffs, and flexible integration with custom workflows with the same enterprise-grade control and compliance. 

More conversational, capable, and configurable 

The Customer Intent Agent’s conversational abilities have been significantly enhanced. Natural language instructions at the organizational and intent level guide the agent’s behavior, enabling more natural, dynamic, and context-aware interactions. The system now supports instructions written in natural language. This allows for dynamic interviews, conditional logic, and seamless transitions between related intents. 

Key improvements: 

• Dynamic interviews: The agent can persist information from self-service, suggest next questions, and handle intent shifts or interruptions gracefully. 

• Instruction-driven resolution: Playbooks provide step-by-step guidance for resolving intents, leveraging both knowledge articles and connected actions. 

• Intent-level mapping: The agent can map intents to knowledge, actions (via connectors), or custom AI agents. This enables tailored resolution logic and flexible integration with proprietary systems. 

• Improved disambiguation: The agent asks clarifying questions and provides immediate feedback. This ensures customer intent is accurately captured, even across thousands of topics. 

Voice support 

The Customer Intent Agent now supports voice for self-service as well as chat. It uses agentic AI techniques to provide a higher quality of self-service and lower chance of escalation by providing a natural flow of conversation, focusing on narrowing down the customer’s reason for calling and identifying the best solution. 

Key improvements: 

• Natural conversation: The agent speaks in a natural empathetic back-and-forth manner, without robotic prompts.  It’s just a conversation that flows, reducing the chance of the call escalating. 

• No pre-programming: The agent dynamically plans the conversation without pre-programmed flows. It reasons over the customer’s speech, narrowing down the intent and identifying the most appropriate solution. 

• Robust speech recognition: The customer speaks naturally, as they would to a person, and the agent understands. 

• Responsive: Fast, natural responses keep the conversation flowing, which is critical in a voice conversation. 

Smarter intent discovery and management 

A core advancement in this release is the overhaul of intent discovery. The Customer Intent Agent now leverages business context, product hierarchies, and industry profiles to generate more relevant, manageable, and actionable intents. Fewer, higher-quality intents make it easier for admins to approve, merge, and segment them by line of business. The new analytics dashboards and semi-autonomous approval flows help organizations quickly identify and promote the most impactful intents. This streamlines the management process and ensures that intent libraries align with real customer needs.  

Key improvements: 

• Business context-aware mining: Intents are now generated with additional context, reducing noise and improving accuracy. 

• Granular control: Admins can filter, approve, or merge intents in bulk, and segment libraries by line of business. 

• Industry and product integration: Mining can be guided by industry and product family, ensuring that intent libraries reflect the organization’s unique landscape. 

• Sample-driven validation: Admins can now view sample cases and conversations that led to intent generation, increasing confidence in the mining process. 

Autonomous and semi-autonomous learning loops 

The July release introduces full and semi-autonomous learning loops. The Customer Intent Agent can continuously improve by extracting new intents and questions from completed conversations. The improved feedback loop, backed by usage data and dashboards, supports the authoring of high-ROI solutions and the creation of knowledge articles. The Customer Intent Agent closes the loop between customer interactions and organizational learning. 

If your organization wants to reduce manual effort, improve resolution times, and deliver consistent customer experiences, now is the time to explore what this agent can do. Preview the Customer Intent Agent in your Dynamics 365 Customer Service environment today! 

Learn more 

To learn more, read the documentation: Autonomous service agents in Dynamics 365 (preview) | Microsoft Learn

The post Automate intent discovery and resolution with Customer Intent Agent appeared first on Microsoft Dynamics 365 Blog.

An update is coming to the closed conversation form in Dynamics 365 Contact Center.  If the administrator has enabled the option, service reps can view the recordings and transcripts at the end of the conversation. They can also view sentiment analysis, key metrics, AI summary, post call survey results, and sessions in the conversation journey. Users can also access the AI summary for voicemails. Note that it will be a summary of the entire transcript (not just the voicemail). 

The Conversation Journey card provides comprehensive information regarding the number of agents in the call and other participants involved. It also provides the duration, date, and time for each segment of the conversation. A new session begins whenever a primary agent is reassigned to the conversation. Each session header displays the start time and total duration. Within each session, the primary agent is identified, including the time they were assigned and their participation status. When the primary agent exits the session, the session concludes, and the system records the closure reason along with the time of occurrence.  

To view Conversation Summaries, enable summaries in Copilot Service admin center by going to Insights and then Summaries. If a summary was generated when the call ended, it will be displayed. Otherwise, a new summary will generate when the transcript viewer loads.  

Admins can enable the Post-conversation survey tab in the conversation form in Copilot Service admin center under Voice call experiences. It displays CSAT scores, textual customer feedback, First Contact Resolution outcomes, and Net Promoter Scores. If admins don’t enable the tab or customers don’t complete the survey, service reps won’t see the tab in the UI. Calls with surveys go to the survey Microsoft Copilot for Service Copilot Bot after they end, not the agent. Specific entities set by Copilot Studio pull the survey data, and if any values are missing, their header does not appear. 

Learn more about AI summary

Watch a quick video introduction.

To learn more about the AI summary for conversation in Dynamics 365 Contact Center, read the documentation: View call recording and transcripts for voice channel | Microsoft Learn

The post Use the Contact Center AI summary for call quality management appeared first on Microsoft Dynamics 365 Blog.

Microsoft Dynamics 365 Customer Service now offers voicemail as part of its omnichannel engagement capabilities. Currently in preview, voicemail can be a valuable component of your call queue management, as an overflow option on a voice queue or when a direct call to an agent goes unanswered.

Group and individual call queue management

When customers choose to leave a voicemail instead of waiting in a queue, the message is routed to the default group voicemail queue. Voicemail from any queue lands in the same voicemail queue, making it a one-stop shop for agents who triage voicemails from customers.

When customers call an agent on a direct inward dialing number and the agent isn’t available, they can choose to leave a voicemail. In this case, the message is routed to the default individual voicemail queue and pushed to the agent associated with the direct inward dialing number.

Both the group and individual voicemail queues start out with no agents assigned to them. Agents who are responsible for managing customer voicemails must be added to the group queue. Agents with direct inward dialing numbers must be added to the individual queue.

The voicemail greeting is an automated message that can be customized at the workstream level. If your call queue management includes a Power Virtual Agents bot, the bot can’t prompt customers to leave voicemail. Instead, the call must be escalated to an agent.

Agent experience

Let’s look at voicemail from the perspective of an agent who’s assigned to the default group voicemail queue. Noticing that a customer has left a voice message, you pick it from the Agent dashboard. Now assigned to you, the voicemail appears in your Activities view. You can also configure voicemails to appear in your Inbox view.

Opening the voicemail form, you can see the customer’s name or number, play back the voicemail, and read a transcript. After you take an action on the voicemail, you can close it. It becomes read-only and moves to your Closed work items view.

Learn more

To learn more about using voicemail as part of your call queue management, read the documentation:

• Use voicemail to manage inbound calls | Microsoft Learn
• Configure the inbox | Microsoft Learn

Not yet using Dynamics 365 Customer Service? Take a tour and get a free trial.

The post Add voicemail to your call queue management appeared first on Microsoft Dynamics 365 Blog.