Every mainframe shop has the same conversation eventually:
“Did you know our systems programmer is retiring in 18 months?”
“No. Why didn’t anyone make a note of that?”
“What do we do now?”
For a long time, the answer was “hire more COBOL programmers,” a strategy that gets harder every year, since the talent pool isn’t growing to match demand. Today, the relevant answer to the crisis of retiring mainframe systems programmers isn’t a hiring plan. Instead, companies that rely on mainframes should use AI that reads code so that their team doesn’t have to reconstruct 30 years of business logic from memory.
IBM watsonx Code Assistant for Z delivers the mainframe coding expertise that companies are lacking due to the widening system programming skills gap.
What Bridging the Systems Programmer Skills Gap Looks Like in Practice
AI tools like IBM’s watsonx Code Assistant for Z don’t replace your systems programmers. They give a new hire (or an overburdened senior one) a way to ask questions about mainframe applications in plain English and get answers. Conversational prompts may include:
- What does this program do?
- What workloads depend on it?
- Why does X variable exist?
When asked these questions, AI provides answers that used to live entirely in one person’s head.
IBM watsonx Code Assistant for Z Use Cases
Watsonx Code Assistant for Z covers a few use cases enterprises are already using AI to fulfill.
Code Explanation
Turning COBOL, JCL, PL/I, and REXX into natural-language documentation that even your newest hire can read
Business Rule Discovery
Surfacing the “why” behind decades-old logic, such as discount calculations, eligibility rules, and compliance checks, without conducting a scavenger hunt through the codebase
Impact Analysis
Mapping what breaks if you touch a given program, before you touch it
None of these processes require ripping out the mainframe or migrating off it. Code Assistant for Z is built to work with the mainframe environment you already have running.
Why the Systems Programmer Skills Gap Matters Now
The skills gap conversation isn’t hypothetical anymore. The mainframe systems programmer skills gap is showing up in survey data across the industry. Now, “skills and workforce” sit alongside “security and modernization” as a top five concern for mainframe shops. The situation has become so dire that some are referring to the problem as a “skills cliff.”
The organizations doing something about the skills cliff aren’t waiting for a perfect hiring market. They’re using AI to compress the time it takes a new employee to become proficient in decades-old systems from years to weeks. According to the 2025 Mainframe Survey report, financial services firms are already using GenAI in mainframe environments to decipher code and uncover problems.
How to Bridge the Mainframe Skills Gap
Mainframe shops with retiring systems programmers have options for overcoming an impending skills gap. Adopting IBM watsonx Code Assistant for Z is one option. However, companies can also outsource systems programming to a trusted provider.
At PSR, we’ve spent years being the outside team that keeps mainframes healthy when the IT bench isn’t as deep as it used to be. We know that using AI to understand coding doesn’t change your company’s relationship with an internal mainframe team. Instead, we ensure that using AI for systems programming makes your team more effective.
If you’re staring down a retirement and wondering whether tribal knowledge will walk out the door with that person, we can help you with Mainframe Systems Programming services or IBM watsonx Assistant for Z Systems. It’s worth having a conversation about the mainframe skills gap before it widens.
Talk to PSR about what’s in your COBOL codebase before you lose the person who knows.