5.1 Meet the Ecosystem: Copilot, Azure AI, Power Platform, Fabric5.1.1 The Infrastructure Foundation (Figure 5-1—Bottom Circle)5.1.2 Data and Orchestration (Figure 5-1—Data Orchestration Circle)5.1.3 AI Service Suite (Figure 5-1—AI Service Suite Circle)Azure AI Agent Service—Managed Runtime and OrchestrationMicrosoft Agent Framework—Unified SDK for Prompt Routing, Memory & Multi-Agent WorkflowsAzure (OpenAI + Cognitive) AI Services/Azure AI Foundry—Models for Language, Vision, Speech, KnowledgeCopilot Studio—No-Code Canvas and Fine-Tune Portal5.1.4 Business Applications Layer (Figure 5-1—Business Applications Circle)Dynamics 365 Agents—Embedded Intelligence in Every RolePower Automate—The Glue That Binds AI to ActionCopilot for Microsoft 365—An Everyday AI Companion5.2 Deciding Which Layer to Use5.2.1 Focus on Mapping Where Your Data Lives to Each Tool’s Connector Framework5.2.2 Not Every Agent Needs to Interact with a User Through Conversation5.2.3 Start at the Top and Drill Deeper5.3 Agentic Framework: Tool Combination Deep Dives5.3.1 Implementing Copilot Studio5.3.2 Azure AI Agent Service Patterns That Will Have You Wiring Agents Together Like A ProStart with Model Context Protocol (MCP)Orchestrating Agent-To-Agent (A2A) HandoffsPicking the Right Memory Store5.3.3 Hardening the Business Applications Layer for AI SuccessPower Automate: The Glue (and the Human Brake Pedal)Dataverse Tables As Shared MemoryApplication Insights for Deep TelemetrySecurity Roles and Data-Residency Controls5.4 Success Stories and Real World Applications
