GM Dynamics & Agentic AI Consulting

This update focuses on moving from “Reactive” service to “Proactive” outreach, using AI to trigger personalized conversations across digital channels before a customer even reaches out for help.

Real-time User Journey

This journey highlights how a company like an airline or retailer uses the feature to prevent service issues:

1. Event Trigger: A flight is delayed by two hours. This system event is captured in Dynamics 365.
2. AI Segment Identification: The AI analyzes the impacted passengers and identifies those with tight connecting flights.
3. Proactive Outreach: Instead of waiting for passengers to call the contact center, the system automatically sends a personalized WhatsApp or SMS message via a Conversational Journey.
4. Natural Language Interaction: The customer replies via WhatsApp. An Autonomous Agent handles the conversation, offering rebooking options or meal vouchers.
5. Seamless Handoff: If the customer has a complex request (e.g., special assistance), the AI agent summarizes the chat and transfers it to a human agent, who sees the full context immediately.

Step-by-Step: How to Enable This Feature

Since this feature is in Public Preview, it must be enabled within a sandbox or preview-enabled environment.

• Step 1: Access Power Platform Admin Center

Sign in to the Power Platform Admin Center.

• Step 2: Enable Preview Features

Select your environment > Settings > Product > Features. Under the “AI and Copilot” section, toggle On the “Proactive Engagement (Preview)” setting.

• Step 3: Configure Dataverse Triggers

Go to the Customer Service Admin Center. Under Engagement, select Proactive Triggers. Define which Dataverse events (like a field update or a new record) should initiate a journey.

• Step 4: Design the Conversational Journey

Open the Customer Insights – Journeys (formerly Marketing) or the Contact Center workspace. Use the “Journey Designer” to drag and drop “Proactive Message” blocks into your workflow.

• Step 5: Link Copilot Studio Agents

For the conversational part, link your Microsoft Copilot Studio bot to the journey so it can handle the natural language replies autonomously.

• Step 6: Test and Publish

Use the “Test Send” feature to verify the SMS/WhatsApp output before publishing the journey to live customers.

Infographic: Proactive vs. Reactive Service

Feature	Reactive Service (Old)	Proactive Engagement (New)
Initiation	Customer starts the contact.	System starts the contact based on data.
Channel	Customer chooses (often high-cost phone).	Contextual channel (SMS, WhatsApp, App).
AI Role	Helps agents answer questions.	Autonomous Agents solve problems before they escalate.
Volume Impact	High spike in inbound calls during crises.	Deflects inbound calls by resolving issues early.
Customer Sentiment	Frustrated/Stressed.	Delighted/Empowered by the early warning.

References

AI-powered proactive engagement and conversational journeys with Microsoft Dynamics 365 in public preview

Improve contact center operations with shift-based routing (Optimizing work distribution by aligning Unified Routing with real-time workforce schedules).

Real-time User Journey

This journey illustrates how shift-based routing prevents customer interactions from “getting lost” when agents log off:

1. Incoming Request: A customer starts a live chat at 4:55 PM regarding an urgent billing issue.
2. Shift Check: The Unified Routing engine doesn’t just check if agents are “Available” (green status); it consults the Shift Schedule.
3. Filtered Selection: Agent A is available but their shift ends at 5:00 PM. Agent B is available and their shift runs until 8:00 PM.
4. Intelligent Assignment: The system bypasses Agent A (to prevent a mid-conversation disconnect or a forced transfer) and routes the chat directly to Agent B.
5. Seamless Resolution: Agent B has ample time to resolve the issue within their scheduled hours. Agent A is able to clock out exactly on time without being “trapped” by a late-arriving work item.
6. Off-Hours Handling: If a request comes in when no shifts are active, the system automatically triggers the “Out of Hours” workflow (e.g., routing to a bot or creating a callback) rather than letting the chat ring in an empty queue.

Step-by-Step: How to Enable This Feature

To transition from standard availability routing to shift-based routing, follow these steps:

• Step 1: Admin Center Access

Sign in to the Customer Service admin center or Contact Center admin center.

• Step 2: Enable Workforce Management

Navigate to Operations > Workforce Management. Ensure that Shift Planning is active and that agent shifts have been populated in the system.

• Step 3: Configure Assignment Methods

Navigate to Routing > Queues. Select a queue and go to the Assignment method section.

• Step 4: Create/Edit Assignment Rule

Select a custom assignment method and go to the Rules tab. Under the Prioritization or Assignment logic, look for the “Check agent availability” setting.

• Step 5: Toggle Shift-Based Routing

Enable the toggle for “Assign work only to agents with active shifts.” * Step 6: Define Buffer Time (Optional)

Set a “buffer period” (e.g., 15 minutes). This ensures that if a chat takes 20 minutes on average, no new chats are assigned to an agent within 15 minutes of their shift ending.

• Step 7: Save and Verify

Save the settings. Use Routing Diagnostics to confirm that work items are being correctly ignored by agents who are “Available” but not currently on a scheduled shift.

Infographic: Benefits of Shift-based Routing

Benefit	Impact on Agents	Impact on Customers
Reduced Transfers	Agents aren’t forced to hand off chats at the end of their shift.	Better continuity; no need to repeat information to a second agent.
Punctual Log-offs	Improves work-life balance by preventing “last-minute” assignments.	Faster resolution as agents aren’t rushing to finish a shift.
Optimized Capacity	Work is only given to those with the time to finish it.	Lower abandonment rates during shift changeovers.
Presence Accuracy	Status is ignored if the agent isn’t officially “on the clock.”	Ensures interactions don’t “stuck-ring” to an agent who forgot to log out.

References

Improve contact center operations with shift-based routing

Intent-based routing transforms customer support with AI (Leveraging Generative AI to automatically detect customer needs and route work items without manual rule-setting).

Real-time User Journey

This journey highlights how AI removes the “guesswork” from traditional keyword-based routing:

1. Customer Inquiry: A customer sends a complex, multi-sentence email: “I received my order #1234 but the screen is cracked and I’d like a refund or a replacement as soon as possible.”
2. AI Intent Detection: The Intent-based Routing engine analyzes the natural language of the email. Instead of just looking for the word “cracked,” it understands the underlying intent is “Product Damage” and “Returns.”
3. Automated Classification: The AI automatically tags the conversation with the correct “Intent” category and assigns a high-priority level because it involves a damaged shipment.
4. Precision Routing: The system identifies that this specific intent requires a “Returns Specialist” with “Hardware” skills. It bypasses general triage and sends the case directly to the specialist’s queue.
5. Agent Resolution: The specialist receives the case already categorized. They don’t need to spend time re-classifying the issue, allowing them to initiate the replacement process immediately.

Step-by-Step: How to Enable This Feature

Intent-based routing is part of the AI-enhanced Unified Routing suite. Follow these steps to configure it:

• Step 1: Admin Center Access

Sign in to the Customer Service admin center or Contact Center admin center.

• Step 2: Enable AI Routing

Navigate to Routing > Work classification. Ensure the toggle for AI-based routing is set to On.

• Step 3: Create Intent Categories

Navigate to Intents under the routing menu. You can either manually create intent categories (e.g., Billing, Technical Support) or allow the AI to suggest categories based on your historical data.

• Step 4: Train the Intent Model

Provide sample phrases or point the system to your existing case data so the AI learns how customers describe specific problems. (Microsoft provides pre-trained models for common service scenarios to speed this up).

• Step 5: Set Up Intent-based Rules

In your Workstream settings, go to Route to queues. Create a rule where the condition is “If Intent Equals [Product Damage]”, then route to the [Returns Queue].

• Step 6: Monitor and Refine

Use the Routing Diagnostics tool to see how the AI is classifying intents. You can “Thumbs Up” or “Thumbs Down” classifications to help the model learn and improve its accuracy over time.

Infographic: Keyword vs. Intent-based Routing

Capability	Keyword-Based (Legacy)	Intent-Based (AI)
Logic	Matches specific words (e.g., “Broken”).	Understands context and meaning.
Maintenance	Requires manual update of thousands of rules.	AI learns and adapts to new phrases.
Accuracy	High risk of “False Positives.”	High precision even with complex language.
Triage	Often requires human manual re-routing.	Fully autonomous classification and routing.
Setup Time	Months of rule-building.	Days of AI training/Pre-built models.

References

Intent-based routing transforms customer support with AI

Email editor template changes in Dynamics 365 Customer Service (Deprecation of legacy editors in favor of a single, unified modern rich text editor).

Real-time User Journey

This journey illustrates how a Customer Service Administrator transitions from legacy templates to the modern editing experience:

1. System Notification: The Admin receives notice that the legacy rich text and enhanced email template editors have been deprecated as of May 23, 2025.
2. Accessing Templates: The Admin navigates to the Email Templates section in the Customer Service hub to update a standard “Case Resolution” template.
3. Automatic Migration: Upon opening the existing template, it automatically loads into the new Modern Rich Text Editor. No manual migration of the content is required.
4. Advanced Styling: The Admin notices the new HTML support. They switch to the HTML view to add specific CSS width units to ensure the template renders perfectly across mobile and desktop devices.
5. Quality Check: The Admin removes a nested table structure that was causing formatting issues in the old editor, taking advantage of the modern editor’s improved reliability.
6. Deployment: After a “Smooth Preview” check, the template is saved and immediately becomes available for agents to use when closing cases.

Step-by-Step: How to Enable This Feature

Because this is a platform-wide deprecation and upgrade, the feature is enabled by default for all users. However, admins should take the following steps to manage the transition:

• Step 1: Verify the Editor

Navigate to Settings > Templates > Email Templates. Open any template; it should now display the modern toolbar and offer an HTML editing tab.

• Step 2: Review Legacy Layouts

Since the “drag-and-drop” layout editor is being phased out, open any templates that used complex layouts to ensure they have converted correctly to the rich text format.

• Step 3: Update HTML Widths

To prevent rendering errors at runtime, Microsoft recommends manually validating that appropriate width units (e.g., width=”100%” or specific pixel values) are present in the HTML code of your templates.

• Step 4: Flatten Table Structures

If a template looks “broken,” check the HTML source for nested tables. The modern editor performs best with simplified table structures.

• Step 5: Admin Center Toggle (If applicable)

In some environments, you may need to go to the Power Platform Admin Center > Environments > [Your Environment] > Settings > Email > Email settings to ensure “Enhanced email template editor” is toggled correctly if you wish to use specific advanced features.

Infographic: The Shift to Modern Email Templates

Feature	Legacy Editor (Deprecating)	Modern Editor (New Standard)
User Interface	Split between “Rich Text” and “Enhanced.”	Unified: One interface for all templates.
Design Method	Drag-and-drop layout blocks.	Rich Text + HTML: Greater precision and branding.
HTML Support	Limited or inconsistent.	Full HTML Editing: Style content exactly as needed.
Reliability	Known formatting and preview issues.	Enhanced Stability: Smoother rendering at runtime.
Legacy Support	Phasing out.	Auto-conversion: Existing templates open here automatically.

References

Email editor template changes in Dynamics 365 Customer Service

Enhance supervisor visibility by monitoring waiting chats (Improving insights and management for asynchronous conversations in Dynamics 365 Contact Center).

Real-time User Journey

This journey illustrates how a supervisor manages “Waiting” conversations to prevent customer abandonment:

1. Asynchronous Shift: An agent is handling a chat regarding a complex insurance claim. They ask the customer to upload documents and move the conversation to a Waiting state.
2. Agent Log-off: The agent’s shift ends while the chat is still in the waiting state.
3. Supervisor Review: The Supervisor opens the Omnichannel Real-time Analytics Dashboard. Unlike before, they can now see a list of conversations currently in the “Waiting” status.
4. Transcript Monitoring: The Supervisor clicks Monitor on a specific waiting chat. They review the live transcript and see how long the customer has been waiting for a follow-up.
5. Intervention: Realizing the original agent is now offline, the Supervisor uses the dashboard to Transfer the waiting chat to an active agent.
6. Seamless Continuity: The new agent receives the chat with full context, ensuring the customer is assisted immediately once they provide their documents, regardless of the original agent’s availability.

Step-by-Step: How to Enable This Feature

As this is an enhancement to the existing Real-time Analytics, follow these steps to ensure visibility is active:

• Step 1: Update to Latest Version

Ensure your Dynamics 365 Customer Service or Contact Center environment is updated to the May 2025 release (part of 2025 Release Wave 1).

• Step 2: Access Supervisor Workspace

Open the Customer Service Workspace or Contact Center Workspace app.

• Step 3: Navigate to Analytics

On the site map, go to Real-time Analytics and select the Ongoing Conversation report.

• Step 4: Configure Report Filters

Ensure that the “Status” filter in your report view includes the Waiting state. This state is typically triggered when an agent uses the “Wait” button or when an automated timeout occurs in async channels.

• Step 5: Assign Permissions

Verify that the user has the Omnichannel Supervisor security role. This role is required to see the “Monitor” and “Assign/Transfer” buttons within the analytics dashboard.

• Step 6: Enable Monitoring for Async Channels

In the Customer Service Admin Center, go to Insights > Real-time analytics settings and ensure that monitoring is enabled for the specific messaging channels (Chat, WhatsApp, etc.) your team uses.

Infographic: Supervisor Visibility Over Waiting Chats

Feature	Old Limitation	New Capability
Visibility	Waiting chats were “invisible” in real-time dashboards.	Full Transparency: View all chats in “Waiting” state.
Actionability	Supervisors couldn’t intervene until a chat was “Active.”	Proactive Monitoring: Review transcripts while in waiting.
Resource Management	Chats could get “stuck” if an agent went offline.	Instant Transfer: Move waiting chats to active agents.
Customer Impact	Long delays and high abandonment.	Faster Response: Continuous support across agent shifts.

References

Enhance supervisor visibility by monitoring waiting chats