Mainframe developers are under pressure to modernize faster, while the number of experienced COBOL developers continues to decline. Many organizations are seeing a growing skills gap, as seasoned experts retire and newer developers step in to maintain and evolve complex, business-critical applications. 

One of the biggest challenges these teams face is understanding existing application logic. Documentation is often incomplete, outdated, or missing altogether. This creates dependencies on subject matter experts (SMEs), slows down onboarding, and makes modernization efforts riskier.

The latest release of IBM watsonx Code Assistant for Z delivers capabilities designed to close this gap: Business rules discovery and explain and documentation generation. Together, they help developers understand applications faster, preserve institutional knowledge, and keep documentation current—all powered by generative AI.

1. Business Rules Discovery: Understand Business Logic Faster

With the business rule discovery capability, developers can rapidly discover and document business rules in COBOL applications using AI to reduce business risk and accelerate modernization while preserving core business logic.

Powered by a next-generation, containerized architecture, watsonx Code Assistant for Z introduces a simplified, web-based interface that integrates AI-powered chat and IBM Z software AI agents. This new experience enables developers and analysts to interact with applications conversationally—asking natural-language questions about code behavior, data flows, and business rules.

This capability helps teams:

• Locate relevant programs instantly: Start by using AI-powered chat to quickly identify which programs are associated with the business process or function of interest.
• Surface key variables automatically: Once a program is selected, watsonx Code Assistant for Z automatically generates a prioritized list of variables—each with an AI-generated description—to help developers pinpoint the variables that drive the desired business logic.
• Generate draft business rules with context: The assistant then produces a detailed draft of the business rule, summarizing its purpose, logic, and key elements. This draft can be reviewed, refined, and exported for validation or integration into modernization tools.

This capability turns previously complex COBOL applications into transparent, explainable assets—unlocking the institutional knowledge needed to modernize with confidence.

2. Code Explanation with documentation generation: keeping application knowledge current

Alongside business rule discovery, watsonx Code Assistant for Z enables AI-driven documentation generation for COBOL applications, with additional languages to be supported in future releases. This feature is currently available as a private preview through November.

These capabilities allow teams to generate natural-language explanations of complex programs and automatically create structured documentation that stays current as applications evolve.

This addresses long-standing pain points:

• Reduce the knowledge gap: Novice developers can quickly understand existing applications, accelerating onboarding and reducing the reliance on SMEs.
• Free up experts: Experienced developers spend less time explaining traditional logic and more time driving modernization initiatives. 
• Facilitate modernization planning: With clearer insights into program intent and dependencies, teams can more effectively assess which components to refactor, rewrite, or retain.

These explanations are contextual and fast, ideal for developers navigating unfamiliar code or troubleshooting production issues. Unlike static documentation, explanations are generated in real time, reflecting the current state of the codebase. Documentation generation transforms the way teams manage knowledge on Z: documentation becomes a living artifact, not a static deliverable.

Preserving Expertise and Enabling Modernization

These new capabilities reinforce IBM’s vision for watsonx as a trusted AI platform for enterprise modernization. By embedding generative AI directly into the Z development lifecycle, watsonx Code Assistant for Z helps organizations bridge the skills gap and de-risk transformation initiatives.

AI doesn’t replace expertise—it amplifies it. It makes institutional knowledge accessible, actionable, and sustainable. Developers can learn faster, teams can modernize more effectively, and organizations can evolve their mainframe applications as part of a broader, cloud-integrated strategy.

Building a Sustainable Future for Z

As AI and automation become core to enterprises, watsonx Code Assistant for Z demonstrates how innovation on Z can keep pace with the modern hybrid cloud world securely, responsibly, and without losing the trusted reliability that makes Z mission critical.

With code understanding and documentation generation, organizations gain more than productivity—they gain confidence in their ability to modernize, sustain, and grow their Z investments for the next generation.

Learn more about how IBM watsonx Code Assistant for Z is helping clients unlock their application knowledge and accelerate their journey to modern, AI-powered development. Explore our interactive demo to get a first look into watsonx Code Assistant for Z’s key capabilities or get in contact with an IBM expert for a live demo.