AI is set to reboot the mainframe

Tomorrow’s market-leading companies are making this critical technology investment today.

Even as technology advances, mainframes remain one of the technological backbones of business. Every day, millions of airline tickets, credit card transactions, and insurance claims are securely processed by these high-performance computers. 

But mainframes don’t—and shouldn’t—run in silos. Ninety-six percent of the 500 respondents in Kyndryl’s 2024 State of Mainframe Modernization Survey Report say that they are moving an average of 36% of their workloads off the mainframe, operating them in a hybrid environment. Meanwhile, 89% say that mainframes remain essential to their business strategies and operations. 

“There’s a reason why mainframes are still around—they run the world’s most mission-critical workloads,” says Petra Goude, global practice leader for Core Enterprise and zCloud at Kyndryl, an IT infrastructure services firm. “But to ensure increased business value, efficiency, and optimized costs, organizations need to modernize their mainframe estates. It’s no longer a question of whether or not to modernize your mainframe. It’s a question of finding the right platform for each workload, based on the outcomes you want to achieve.”

The rise of AI is transforming the IT landscape and spurring the need for innovation and agility in a modern hybrid cloud: a mixed computing environment where applications are run using a combination of public and private clouds and on-premises computers. As a critical component of this modern hybrid cloud, mainframes are key to meeting this demand.

Interdependence of AI and the mainframe

As business leaders increasingly adopt a hybrid IT approach, accelerating the mainframe application, data, and infrastructure transformation will be key to increasing business value. This is one of the areas where businesses face a skills shortage and where AI and generative AI can help.

“AI is one of the most disruptive technologies that we’ve seen in terms of mainframe modernization,” says Richard Baird, senior vice president and global chief technology officer, Core Enterprise and zCloud for Kyndryl.

Because many mainframe applications don’t have thorough or up-to-date documentation, generative AI can help users understand them by representing what the application is doing in natural language. This then helps engineers figure out what they need to do with the application to modernize it. And that’s just one example.

In today’s AI era, businesses are also looking to get the most out of the large volumes of data that mainframes contain by using it to train new models. For example, insurance companies can use complex data analytics to help actuaries provide better insight to insurance agents. Often, organizations can enable more powerful uses of their own data by updating their mainframe hardware and software to be able to access data from external components, such as the cloud.  

“Mainframe applications and data contain a lot of insights into how your business actually operates, so they need to be understood and modernized,” says Baird. “Just like if you run your car for many years, sometimes it needs a service to understand that a new set of tires are required to get from point A to point B. Modernization is the same—a good investment to keep the business going.” 

Kyndryl helps organizations unlock the full potential of their mainframe applications and data assets with AI. The company’s solutions can help ingest mainframe data to build custom AI models, extend new or existing applications to use those AI models, and even exploit AI capabilities in the operating systems to provide faster and more efficient runtimes. “Mainframe modernization is required for the business use of AI, and AI can enable the transformation,” says Goude.

Today, applications rarely operate or rely on data from a single platform. But 85% of survey respondents say they are lacking an integrated view of the entirety of their operational data, with 92% saying it’s an important need. With this in mind, Kyndryl Bridge, an AI-infused open integration platform, is built to extend AI capabilities across the hybrid cloud to provide operational insights and facilitate the deployment of responsible AI-based solutions. These expanded AI capabilities are supported and enabled by Kyndryl Consult, the company’s advisory and implementation services arm built on decades of experience in mainframe technologies. 

Meeting the AI talent demand

Overall, embedding AI into mainframe and hybrid cloud environments can enhance business outcomes, augment human capabilities, streamline automation of business processes, and generate actionable insights from data. However, utilizing responsible AI and generative AI in mainframe environments requires a thorough understanding of the technology, in addition to cloud and mainframe domain expertise. To address a significant skills gap that many organizations face, Kyndryl is training more than 5,000 of its mainframe professionals in AI and generative AI skills, augmenting their already extensive certifications and decades of experience in mainframe and cloud technologies. 

“Most of these applications were written 10, 20, 30, and even 50 years ago, so it’s challenging to find people who are skilled in mainframe computing and understand other platforms and how to carry out integrations,” says Goude. “Kyndryl has the global skills, expertise, and experience across multiple industries to help with that.”

While AI promises to be a valuable tool to accelerate mainframe modernization, the potential for the technology will continue to evolve. Still, now is the time to get started. “Don’t wait and keep the status quo,” says Goude. “If you don’t have the skill set, work with partners who understand how the hybrid cloud infrastructure works together with these mainframe systems. The world’s most mission-critical workloads run on mainframes—make sure you continue to future-proof them and grow your business.”