Build advanced agents in Copilot Studio by extending beyond basic conversational flows and native knowledge sources. Attendees will learn how to design agents that interact with tools, manage context across steps, and retrieve information from structured and unstructured data sources.

Through guided, hands-on exercises, participants will build agents that incorporate Model Context Protocol (MCP) for tool interaction and context management, Azure AI Search for controlled and scalable information retrieval, tool discovery and invocation, agent and tool chaining, and contextual handoff across agent actions.

Along the way, we examine how Copilot Studio communicates with external tools, how context is passed and reused, and how search results are consumed and applied by agents. The workshop emphasizes practical design decisions that affect agent behavior, reliability, and maintainability as solutions grow.