contact-centre, Customer-service

Bringing end-to-end visibility with screen recording in Dynamics 365 Contact Centre

Bringing end-to-end visibility with screen recording in Dynamics 365

(A unified capability for Dynamics 365 Customer Service and Dynamics 365 Contact Center that captures an agent’s on-screen workflow to provide context beyond just voice and text.)

Real-time User Journey

This journey illustrates how screen recording helps resolve a complex “process” dispute where a transcript alone wasn’t enough:

  1. Complex Interaction: An agent assists a customer with a high-value insurance claim. The interaction involves navigating three different internal legacy systems and a knowledge base.
  2. Triggered Recording: As soon as the agent accepts the call in the Contact Center, the screen recording starts automatically (or is manually toggled by the service rep).
  3. Visual Workflow: The agent hits a snag in the legacy system, leading to a delay. They search for a workaround in the knowledge base, which the recording captures.
  4. Secure Upload: Upon ending the session, the recording is automatically and securely uploaded to Microsoft Dataverse, linked directly to the specific conversation and case record.
  5. Quality Evaluation: A supervisor reviews the “Critical Failure” flagged by the Quality Evaluation Agent. While the transcript sounds professional, the Screen Recording shows the agent bypassed a mandatory compliance checkbox in the internal system.
  6. Targeted Coaching: The supervisor uses the video to show the agent exactly where they deviated from the process, using the recording as a visual training tool to prevent future errors.

Step-by-Step: How to Enable This Feature

Administrators can enable and govern this feature within the admin center settings:

  • Step 1: Admin Center Access

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

  • Step 2: Navigate to Recording Settings

Go to Operations > Insights > Call and Screen Recording.

  • Step 3: Enable Screen Recording Controls

Toggle the “Enable screen recording” switch to On. You can choose to enable this for “Voice Conversations” (automatic) and “Other Workloads” (manual toggle for agents).

  • Step 4: Configure Storage & Governance

Ensure your Dataverse storage is configured to handle video files. Set the Retention Policy (e.g., “Delete screen recordings after 90 days”) to align with your regional compliance laws.

  • Step 5: Set Role-Based Access (RBAC)

In the Security Roles section, assign the “Screen Recording Viewer” or “Manager” roles only to authorized supervisors and compliance officers. This ensures that sensitive on-screen data is protected.

  • Step 6: Deploy to Workspace

Go to Agent Experience Profiles and ensure the “Screen Recording” component is visible in the agent’s live conversation widget or productivity pane.

Infographic: Closing the Visibility Gap

AspectTranscript / Audio OnlyWith Screen Recording
VisibilityCaptures “What was said.”Captures “What was done.”
ComplianceHard to verify if policies were clicked.Visual proof of process adherence.
System UsageBlind to technical glitches.Identifies UI/UX bottlenecks in apps.
CoachingBased on verbal cues.Based on actual navigation and speed.
SecurityRedaction limited to text/audio.Role-based secure access to video files.

References

copilot-studio

Securing the Agentic Frontier: Addressing OWASP Top 10 Risks in Agentic AI with Microsoft Copilot Studio

This topic focuses on the governance and security framework required to protect “Autonomous Agents”—which have the power to act on behalf of users—against emerging threats like prompt injection, data exfiltration, and unauthorized tool use.

The 10 failure modes OWASP sees in agentic systems

  1. Agent goal hijack (ASI01): Redirecting an agent’s goals or plans through injected instructions or poisoned content.
  2. Tool misuse and exploitation (ASI02): Misusing legitimate tools through unsafe chaining, ambiguous instructions, or manipulated tool outputs.
  3. Identity and privilege abuse (ASI03): Exploiting delegated trust, inherited credentials, or role chains to gain unauthorized access or actions.
  4. Agentic supply chain vulnerabilities (ASI04): Compromised or tampered third-party agents, tools, plugins, registries, or update channels.
  5. Unexpected code execution (ASI05): Turning agent-generated or agent-invoked code into unintended execution, compromise, or escape.
  6. Memory and context poisoning (ASI06): Corrupting stored context (memory, embeddings, RAG stores) to bias future reasoning and actions.
  7. Insecure inter-agent communication (ASI07): Spoofing, intercepting, or manipulating agent-to-agent messages due to weak authentication or integrity checks.
  8. Cascading failures (ASI08): A single fault propagating across agents, tools, and workflows into system-wide impact.
  9. Human–agent trust exploitation (ASI09): Abusing user trust and authority bias to get unsafe approvals or extract sensitive information.
  10. Rogue agents (ASI10): Agents drifting or being compromised in ways that cause harmful behavior beyond intended scope.

Real-time User Journey: Secure Autonomous Execution

This journey illustrates how Copilot Studio’s security layers prevent an “Indirect Prompt Injection” attack:

  1. The Trigger: An autonomous agent is tasked with summarizing a set of incoming emails and syncing action items to a CRM.
  2. The Threat: One of the emails contains hidden malicious instructions (an “Indirect Prompt Injection”) designed to trick the agent into sending sensitive company data to an external personal email address.
  3. Real-time Interception: Before the agent executes the “Send Email” tool, the Microsoft Defender for Agents layer inspects the intent. It identifies that the destination address is not on the organization’s “Allow List” and that the payload contains sensitive keywords.
  4. Governance Block: The agent’s Managed Identity permissions are checked. The system realizes the agent is attempting an action (external exfiltration) that exceeds its scoped authority.
  5. Safe Resolution: The action is blocked. The user (and IT admin) receives a notification that a suspicious activity was intercepted, and the agent continues with other safe tasks.

Step-by-Step: How to Enable Security Features

To align your agents with the OWASP security recommendations using Copilot Studio tools:

  • Step 1: Assign a Managed Identity: Navigate to the agent settings in Copilot Studio and enable Microsoft Entra Agent ID. This ensures the agent has its own identity and doesn’t “ghost” as a high-privilege human user.
  • Step 2: Configure Content Safety: Under Settings > Security, enable Microsoft Azure AI Content Safety. Adjust the sliders to “High” for categories like Jailbreak detection and Protected Material.
  • Step 3: Define Tool Guardrails: In the Tools tab, for every connector (like SAP or Salesforce), set “User Confirmation” to “Required” for sensitive actions (e.g., deleting records or making payments).
  • Step 4: Enable Network Isolation: In the Power Platform Admin Center, configure Virtual Network (VNet) support for your environment to ensure agent traffic never leaves your private network.
  • Step 5: Monitor via Defender: Connect your agent logs to the Microsoft Defender for Cloud dashboard to receive real-time alerts on prompt injection attempts.

Infographic: OWASP Top 10 vs. Copilot Studio Protections

This table summarizes how Microsoft’s platform mitigates the most critical risks identified for LLM agents:

OWASP Risk CategoryCopilot Studio / Microsoft Security Solution
Prompt InjectionDefender for Agents: Scans inputs for malicious “jailbreak” patterns.
Insecure Output HandlingAzure AI Content Safety: Sanitizes agent responses before the user sees them.
Excessive AgencyScoped Managed Identities: Limits what an agent can do based on “Least Privilege.”
Data ExfiltrationDLP (Data Loss Prevention) Policies: Blocks sensitive data from being sent to unapproved domains.
Insecure Knowledge AccessTenant Graph Grounding: Respects existing SharePoint/OneDrive permissions automatically.

References

Customer-service

Microsoft named a Leader in The Forrester Wave™: Customer Service Solutions, Q1 2026

Microsoft named a Leader in The Forrester Wave™: Customer Service Solutions, Q1 2026

(Recognition of Microsoft’s “Agentic” service vision, highlighting the strength of Dynamics 365 Contact Center and Customer Service in modernizing the support experience through autonomous AI.)

Real-time User Journey: The Integrated Agentic Experience

The Forrester report emphasized “horizontal” integration—where AI works across the entire company. This journey reflects that capability:

  1. Smart Recognition: A customer with a “Premium” status calls support. The Unified Routing engine identifies their high-value status and their previous interaction in Microsoft Fabric regarding a delayed order.
  2. Contextual Handoff: Instead of asking the customer to repeat themselves, the AI agent greets them with: “I see you’re calling about your pending order from yesterday. I’ve already checked the shipping status.”
  3. Collaborative Resolution: The issue requires a back-office warehouse update. The agent uses the Microsoft Teams Swarming feature. A warehouse expert is pulled into a sidebar chat automatically, provides the update, and leaves.
  4. Copilot Sidekick: While the agent talks, Copilot generates a “Resolution Draft” based on the warehouse update.
  5. Autonomous Post-Call: Once the call ends, the Case Management Agent autonomously updates the CRM, sends a confirmation email, and triggers a refund in the finance system.
  6. Quality Oversight: The Quality Evaluation Agent scores the call for “Empathy” and “Efficiency” within seconds, feeding the data into the supervisor’s leadership dashboard.

Step-by-Step: How to Enable the “Leader” Capabilities

To implement the specific features cited by Forrester as market-leading, follow these steps:

  • Step 1: Unify Data via Dataverse & Fabric

Ensure your customer data is not siloed. In the Power Platform Admin Center, enable the Microsoft Fabric Link for your Dataverse environment to give AI a 360-degree view of the customer.

  • Step 2: Activate the Agentic Suite

Navigate to the Agent Hub in Dynamics 365 Customer Service. Enable the Case Management Agent and Knowledge Management Agent to move from “assisted” to “autonomous” service.

  • Step 3: Set Up Unified Routing

Go to Customer Support > Routing. Configure “Intents” so the AI can predict why a customer is calling and route them to the best-fit agent or digital bot.

  • Step 4: Enable Copilot for Service

In the Admin Center, go to Agent Experience > Productivity. Toggle on “Copilot Summaries” and “Generative Answers” (connecting it to your internal SharePoint/Knowledge Base).

  • Step 5: Integrate Teams Collaboration

Enable “Embedded Chat using Microsoft Teams” in the settings. This allows agents to “swarm” complex cases with experts from across the organization without leaving the CRM.

  • Step 6: Monitor via AI Evaluation

Deploy the AI Performance Dashboard to monitor the “Three Pillars” (Understand, Reason, Respond) that Forrester highlighted as key to Microsoft’s high score in technical strategy.

Infographic: Why Microsoft was Named a Leader

Forrester Evaluation CriteriaMicrosoft’s Strategic Advantage
Agentic InnovationMoving from chatbots to autonomous Case & Quality Agents.
Platform SynergyDeep integration with Teams, Outlook, and Azure.
Data StrategyReal-time insights via Microsoft Fabric and Dataverse.
Developer ToolsEmpowering IT and “Citizen Developers” via Copilot Studio.
Security & TrustBuilt-in Sensitive Data Redaction and enterprise-grade privacy.

References

copilot-studio, Power Apps

2026 Release Wave 1: Transitioning to the Era of AI-Powered, Agentic Business Applications

This release wave (covering April 2026 to September 2026) marks a foundational shift for Dynamics 365, Power Platform, and Copilot Studio, moving from assistive AI to autonomous agentic workflows that unify data and automate complex processes across sales, service, finance, and supply chain.

Real-time User Journey: The Scheduling Operations Agent (Field Service)

One of the highlighted journeys in this wave involves the Scheduling Operations Agent in Dynamics 365 Field Service:

  1. Event Trigger: A high-priority emergency repair request comes in via a customer portal while the human dispatcher is busy.
  2. Autonomous Analysis: The Scheduling Agent automatically scans all active technicians’ locations, skill sets, and current workloads.
  3. Conflict Resolution: The agent identifies that the best-suited technician is currently on a low-priority maintenance call. It automatically notifies the maintenance customer of a slight delay and reassigns the emergency ticket.
  4. Technician Guidance: The technician receives a real-time update on their mobile device with optimized routing and a summary of the emergency.
  5. Dispatcher Oversight: The dispatcher is presented with a “completed action summary” rather than having to manually drag and drop schedules, moving from “scheduler” to “supervisor.”

Step-by-Step: How to Enable Wave 1 Features

Note: Release Wave features are typically rolled out in phases. Admins can manage them via the Power Platform Admin Center.

  • Step 1: Access the Release Planner: Go to the Microsoft Release Planner to identify which specific features are available for “Early Access.”
  • Step 2: Enable Early Access: Log into the Power Platform Admin Center, select your environment, and under Updates, click “Manage” to opt-in to the 2026 Release Wave 1 early access features.
  • Step 3: Configure Agent Builder: In Copilot Studio, use the Agent Builder to customize role-based agents (like the Sales or Finance Agent) with your specific organizational data.
  • Step 4: Connect to Work IQ: Enable the Work IQ integration within Dataverse to allow your agents to learn from organizational patterns and provide more grounded decisions.
  • Step 5: Deploy via Managed Environments: Use the refreshed Governance and Administration tools to set pay-as-you-go (PAYG) caps on Copilot credits before rolling out to the entire tenant.

Infographic: 2026 Wave 1 Innovation Pillars

The release wave is structured around four strategic areas:

PillarKey HighlightBusiness Value
Agentic ERPAutonomous Sales/Purchase agents in Business Central.Reduces manual data entry and accelerates procurement cycles.
Unified DataCustomer Insights as the “grounding layer” for AI.Ensures agents make decisions based on real-time, 360-degree customer views.
Low-Code PortalsSecurity Agent for Power Pages.Allows non-developers to build secure, AI-integrated customer portals.
Daily Command CentersFinance & Sales Agents in M365 Copilot.Brings ERP data directly into Excel, Outlook, and Teams for faster analysis.

References

contact-centre, Customer-service

2026 Release Wave 1: Transforming Business Processes with Agentic AI in Dynamics 365

2026 Release Wave 1: Transforming Business Processes with Agentic AI

(A comprehensive roadmap detailing hundreds of new capabilities across Dynamics 365, Power Platform, and Copilot Studio, focusing on autonomous agents and “human-in-the-loop” productivity.)

Real-time User Journey: The Autonomous Supply Chain

While the release covers many areas, the “Autonomous Procurement” journey is a primary highlight of this wave:

  1. Inventory Anomaly: An Autonomous Supply Chain Agent identifies a predicted shortage of a critical component due to a regional weather event.
  2. Market Research: The agent proactively researches alternative suppliers in the Microsoft Dataverse and via the web, comparing lead times and carbon footprints.
  3. Drafting the Plan: The agent drafts a procurement plan, including a pre-filled purchase order for a new supplier that meets all sustainability criteria.
  4. Human-in-the-loop Approval: The Procurement Manager receives a notification in Microsoft Teams. They open the Copilot sidecar, review the agent’s reasoning, and approve the order with one click.
  5. Execution & Tracking: The agent submits the order, updates the ERP, and begins tracking the shipment, providing the manager with real-time risk updates as the weather event progresses.
  6. Resolution: The factory avoids a production halt, and the manager spends their day on strategic supplier relationships rather than manual order tracking.

Step-by-Step: How to Enable Wave 1 Features

Release wave features are rolled out in stages. Here is how to manage and enable them:

  • Step 1: Access the Release Planner

Visit the Microsoft Release Planner to see the specific dates for each feature (General Availability vs. Public Preview).

  • Step 2: Enable Early Access

Sign in to the Power Platform Admin Center. Go to Environments > [Your Environment] > Settings > Updates. Click “Manage” and select “Update now” to enable the 2026 Wave 1 early access features in a sandbox environment.

  • Step 3: Configure New Autonomous Agents

Navigate to the Agent Hub within Dynamics 365 or Copilot Studio. Locate the new pre-built agents (e.g., Sales Research Agent or Case Management Agent) and toggle them to Enabled.

  • Step 4: Update App Profiles

Go to Agent Experience Profiles and ensure the new Copilot Pane and Multi-session companion features are added to your users’ workspaces.

  • Step 5: Define Knowledge Sources

In Copilot Studio, point your agents to the updated Microsoft Fabric or Dataverse connectors to ensure they are using your latest enterprise data for their “Agentic” reasoning.

  • Step 6: Monitor Performance

Use the newly released AI Evaluation Dashboards to measure the accuracy and ROI of the Wave 1 features as they are rolled out to production.

Infographic: 2026 Release Wave 1 Pillars

PillarKey FocusFeatured Capabilities
Agentic AIMoving from “assist” to “act.”Autonomous Sales, Service, and Supply Chain Agents.
Unified DataBreaking silos with Dataverse + Fabric.Real-time insights, zero-copy data sharing, and AI “Memory.”
Copilot StudioEmpowering makers to build agents.Visual agent designer, advanced orchestration, and safety guardrails.
Integrated WorkspaceEmbedding AI where you work.Native AI in Teams, Outlook, and the new Desktop Companion App.
GovernanceTrustworthy and secure AI.AI Evaluation frameworks, sensitive data redaction, and audit logs.

References