Skip to main content

Every day, billions of people worldwide book airline and train tickets, withdraw money from ATMs and buy goods and services online, all thanks to the trusty mainframe. 

The fact that mainframe technology turned 60 in April 2024 and still powers so much of modern living is something to celebrate. However, many experienced mainframe engineers are retiring, and the employees replacing them are more proficient in newer coding languages like Python and Java than classic languages like COBOL and PL/I, which underpin 95% of ATM transactions.1

The mainframe skills gap occurs as more companies move workloads to cloud platforms. Yet organizations aren’t choosing the cloud exclusively over the mainframe or vice versa. Most use—and are committed for the long term to—a hybrid cloud approach

While there’s no shortage of cloud computing advocates, we need similar support from technology leaders who recognize that the mainframe, and hybrid clouds, will be part of their core business for years to come. These same leaders can call for solutions to the critical mainframe skills gap. Here are five ways to do both.

Many mainframe engineers are retiring, and the employees replacing them are more proficient in coding languages like Python and Java than COBOL and PL/I.

1. Sponsor classroom learning

Simply put, the world needs more people with mainframe skills. Ideally, education would start in the primary grades and continue through secondary school as a core component of IT skills curricula. Organizations could also partner with universities and other higher education institutions to provide mainframe technology training and sponsor related educational opportunities. 

Mainframe engineers will continue to be valued, as reflected in median salaries.2 Demand for COBOL experts should remain high, even as the need for programmers in RPG, REXX, PL/I, JCL and other languages increases in parallel.1

2. Invest in internal training

Current employees can be a significant source of talent, so encourage staff engineers to learn older programming languages typically used in mainframe environments. It’s more cost-effective to train existing team members than to identify, recruit, hire and onboard external candidates skilled in mainframe computing. 

Generally, reskilling current employees pays for itself in the form of more engaged and productive workers and less attrition.3 However, when linked to larger business objectives or digital transformation initiatives, reskilling programs can impact ROI more dramatically and help to future-proof your organization.4

3. Drive adoption of generative AI

While generative AI holds promise for companies facing mainframe skills shortages, more use cases are needed before organizations can deploy the technology extensively. 

For instance, knowledgeable COBOL programmers can use generative AI to correct code syntax or generate more efficient source code. As the technology evolves, more experts would be able to use generative AI to modernize COBOL code and translate it into Java or other mainframe programming languages.

Current employees can be a significant source of talent, so encourage team members to train other engineers in programming languages often used in mainframe environments.

4. Tap into advanced technologies and processes

Advanced technologies and processes can make working on mainframes more exciting for talented developers who prefer newer platforms. For example, companies can use AI and machine learning on mainframes to gather real-time operational insights to analyze existing applications, identify anomalies, and improve customer personalization.

DevSecOps, the unification of application development lifecycles, and a culture of collaboration can also broaden access to mainframe technology. This would happen as the application development processes between those developing for the mainframe and those developing for cloud environments become more united. In some cases, you could also unify the tools being used.

Meanwhile, organizations can use API approaches, which have been around for decades, to connect applications written in older programming languages to newer programming languages. COBOL applications, for instance, are easy to integrate with Java or Python applications. Combining programming languages is an excellent way for more engineers to begin working with mainframes while upskilling on COBOL.

5. Demand specifics from your partner

IT modernizations can siphon staff from their core duties and require expertise they don’t have. The skills gap created by this talent dispersion prompts 74% of businesses to partner with third parties for complex IT transformations. 

If you’re considering outsourcing all or part of your modernization, search for a partner to address temporary skills challenges and help reskill your current employees. Be sure to ask potential partners:

  • What’s your approach to mainframe modernization?
  • How does it differ from competitors’ methods?
  • Do you provide skills augmentation services to solve immediate issues?
  • Will you accept rebadged employees to maintain flexibility while solving challenges?
  • Do you offer application management services?  

Third-party firms that answer these questions to your satisfaction are more likely to help your company overcome short- and long-term skills gaps and accelerate its digital transformation.

The call for mainframe developers

Thanks to their durability, mainframes are—and will remain—integral to a modern, hybrid cloud approach to digital transformation. So, companies that address the mainframe skills gap should reap the benefits of performance and reliability for years to come. 

Richard Baird is  Senior Vice President and CTO for Kyndryl’s Core Enterprise and zCloud practice.