I recently attended IBM Dev Day: AI Demystified in January 2026, and one topic that really stood out to me was Agentic AI for IBM Z.

When I first started learning about IBM Z, I found myself jumping between multiple documents to understand how things worked. It was a time-consuming process — manually searching through documentation for answers. That’s exactly where Retrieval-Augmented Generation (RAG) can make a difference. With RAG, AI can quickly retrieve relevant information from trusted documents and provide context-aware answers.

During the session, I learned about the introduction of a Z RAG Agent, which is grounded in a vast collection of IBM Z resources — from official documentation to white papers. For any RAG-based solution, success depends on the quality and completeness of the knowledge it’s built upon. Since the Z RAG Agent is trained on IBM Z–specific data, it can efficiently assist both new and experienced users, reducing the need to manually browse through multiple reference materials.

Another exciting development is the Z Upgrade Agent, which can translate generic installation and update commands into mainframe-specific processes using SMPE (System Modification Program Extended) and even execute those operations automatically. There’s also a Z Support Agent capable of analyzing system dumps and assisting with troubleshooting.

These are just a few of the Agentic AI capabilities announced during the session. I’m looking forward to seeing them in action and exploring how they can simplify tasks and transform the IBM Z experience.