Explore the architecture of the Model Context Protocol (MCP) and understand how it enables AI applications and agents to connect with external tools, services, and data sources in this session from the Software Architecture Conference 2026, presented by Rodrigo Díaz Concha.

The session focuses on the architectural foundations of MCP, including how its components and primitives work together to provide a standardized way for AI applications to access tools and resources. Understanding this architecture is important for developers and architects building agentic AI applications, where agents need controlled and structured access to external capabilities and information.

The discussion is particularly relevant to AI engineers, software architects, .NET developers, solution architects, and developers building AI agents and LLM-powered applications. It provides architectural context for understanding how MCP can connect AI applications with external systems while supporting a more consistent approach to tool and resource integration.

MCP is increasingly used as an architectural approach for exposing tools and resources to AI agents in a centralized and controlled manner. Microsoft's MCP learning resources similarly describe MCP in terms of connecting AI applications with tools and external data sources, with applications across MCP, .NET, generative AI, and agentic AI.

Watch the full session to understand MCP architecture, AI agent connectivity, MCP tools and resources, protocol components, and the role of Model Context Protocol in modern agentic AI systems.