Session Evaluation

* AI can be like having an expert assistant working for you 24/7. Discover how agentic AI is transforming z/OS development workflows by acting as your companion that understands mainframe code and business logic. This session introduces AI agents that can autonomously navigate COBOL, PL/I, JCL and other codebases to answer questions, suggest improvements, and help developers understand complex legacy systems. Learn how these AI agents differ from traditional code analysis tools by providing contextual understanding and interactive assistance.

* We'll demonstrate practical use cases including code documentation generation, impact analysis for changes, and automated code review suggestions. Live examples include AI agents helping developers understand unfamiliar programs, trace data flows, and identify potential issues—all without requiring deep expertise in every corner of the codebase. Perfect for teams looking to accelerate onboarding and improve productivity.

Session Evaluation

* AI erupted into almost every aspect of humanity a couple of years ago. Along with overarching concerns that it could eliminate entire categories of jobs, there are more nuanced questions of how good (or bad) AI actually is. Sycophantic behavior, hallucinations, doom loop, and other negative characteristics of AI can harm its usefulness.

* In this session we will discuss general and particular areas of zos application development where AI deficiencies might manifest.

* Utilizing real life examples we will propose strategies to mitigate and overcome these deficiencies, and improve AI’s ability in assisting zos application development and maintenance.