Why mainframes still matter in an AI-driven world

By Elitsa Popova, Technical Consultant, Vertali

Five years ago, I started my career as a Mainframer in Training. This gave me a solid foundation in a supportive environment where I learned the basics and built confidence. Prior to that, I had an internship as a product developer in software, and an apprenticeship as a full-stack web application developer. 

Moving to Vertali marked a significant change. The work here is far more varied, with many different clients and higher expectation – which has pushed me to grow faster. If my earlier experience was about understanding the “well”, working at Vertali feels more like being in the “ocean” – dealing with bigger challenges and gaining a wider perspective.

I’m from Bulgaria, where mentoring and professional development are quite structured, and focused on building strong technical foundations. Overall, I think the UK puts more emphasis on adaptability, communication, and real-world experience. In UK academia – I studied computer, science and software engineering at the University of Bedfordshire – the focus tends to be more academic and theory based. At a company like Vertali, the approach is more practical. Your day-to-day work encourages independence and learning by doing. 

As specialist in z/OS Db2 DBA, I focus on keeping customer database environments stable, secure, and running efficiently for business-critical systems. I work on schema changes and migrations using tools like BMC AMI Change and Catalog Manager, making sure everything is consistent across environments, keeping risks low. 

My work involves impact analysis, performance tuning, and making sure infrastructure changes are managed securely. Because I also have a background in full-stack development, I can understand both the application and the database side, which helps me to deliver more reliable solutions.

Why mainframes still matter

Mainframes continue to play a key role in IT strategies for several reasons. They can manage vast numbers of transactions reliably and securely. They offer extremely high uptime and strong built-in security features. And many organizations rely on systems that have been built over decades, and moving everything to the cloud can be risky and expensive. 

As a result, far from being replaced, mainframes are evolving with modern technologies. AI is built directly into processors like IBM Telum II, allowing real-time decisions during transactions. There is also greater integration with cloud through hybrid models, combining mainframe systems with cloud-based applications. GenAI is helping modernise older systems, for example moving from COBOL to more modern languages. At the same time, tools like Visual Studio Code and GitHub are making development easier, while security is improving with innovative approaches like quantum-safe encryption. 

What lies ahead?

AI represents one of the biggest developments in mainframes, as in so many other areas. In the next few years, AI will become a much bigger part of how mainframes operate. It will help people to make decisions in real time, simplify working with older systems, and automate operations. We are likely to see more self-managing systems through AIOps, with AI handling routine tasks while people focus on higher level decisions. Overall, AI will help improve speed, reduce complexity, and make systems easier to manage.

Elitsa is a Technical Consultant with Vertali, specializing in z/OS Db2 DBA. Co-chair of GS-UK’s WAVEZ 101 working group and a Mainframe Society Ambassador, she was named an IBM Champion in 2025 and 2026.