Stream: Chapel
Time: 10:00 - 10:45
Model Context Protocol (MCP) is rapidly becoming a standard way to connect AI assistants with enterprise systems and tools. But what does it actually take to build an MCP server, and how can it unlock the value of mainframe assets? In this session, we'll explore the fundamentals of MCP, how MCP servers are built, and the patterns used to expose enterprise capabilities to AI applications. Through practical demonstrations, we'll show how MCP can integrate with IBM Z environments and services such as z/OSMF, SDSF, and other mainframe interfaces. Attendees will learn how to create their own MCP server, expose mainframe operations as AI-consumable tools, and enable developers to interact with enterprise systems using natural language. We'll also discuss lessons learned, common challenges, security considerations, and opportunities for innovation at the intersection of AI and the mainframe.
There is currently no attachment for Having Fun with MCP: Building AI-Powered Mainframe Integrations
As a seasoned developer entrenched in the realm of mainframe technologies, my journey began at banking sector, where I honed my craft and developed a profound passion for this intricate domain. Transitioning from the banking sector to academia, I now bring real-world experience to the classroom as a Mainframe Technologies instructor at a Vilnius University While delving into the intricacies of software development for mainframe clients at Rocket Software, I remain deeply committed to fostering the next generation of tech enthusiasts. My dual roles as a developer and educator intertwine seamlessly, empowering me to not only create innovative solutions but also impart crucial knowledge to aspiring professionals.