As AI agents become more capable and autonomous, developers face new challenges around orchestration, memory, and context management. Tools like LangGraph and the emerging Model Context Protocol (MCP) offer a structured way to design multi-step, stateful agents that are easier to build, debug, and deploy. For AI engineers, understanding these new tools is key to staying at the forefront of agent development.
In this code-along, Michael Jesmin, a Generative AI Architect at Capgemini, will teach you how to create an AI agent using LangGraph and MCP. You’ll explore how the Model Context Protocol works, what problems it solves, and how to spin up your own MCP server. Whether you’re experimenting with agents or building production-grade workflows, this session will give you the practical foundation you need.
Presenter Bio
Michael JesminGenerative AI Architect at Capgemini Engineering
Michael is a seasoned engineering leader with 20 years of experience in leveraging cutting-edge AI technologies to drive digital transformation. As a Generative AI Architect at Capgemini Engineering, he develops and manages AI solutions for clients. Previously, Michael was a senior software engineering manager with stints at Dematic, Leuze Electronic, and BAE.
