How banks can achieve a modern core system with fit-for-purpose platforms

By Jonathan Cavell

My bank’s mobile app is easily one of my most used. I make deposits, send payments, split complex bills, and do other tasks quickly, when and where I need to.

The speed and user-friendliness of the app may suggest that all the bank's software is created on cutting-edge technology with modern best practices. But I'd be surprised if this is the reality. In the race to meet customer expectations, banks have tried to deliver new features, applications, and capabilities without modernizing the heart of the bank — the core banking system.

Instead, many of these new applications have been patched on, making the core more difficult to maintain and making changes more complex to orchestrate. At the same time, many banks have allowed significant technology debt to grow while much of the talent who built and understood those systems have retired. The result is a brittle core that must be modernized to accelerate the pace of development and innovation required in our banking industry today.

Banking leaders are well aware of the challenges facing their current core systems—and of the benefits of modernization. In fact, 36% of banks in the U.S. are prioritizing the modernization of internal IT systems that will support a transformation of their core banking system over the next two years.¹

These leaders are also aware of the risks associated with modernization. In a widely-discussed study, Boston Consulting Group found that 70% of digital transformations fall short of their objectives, often with profound consequences.² (When BCG uses words like “profound,” we should all be nervous.)

So, how do we get out from between the need-to-transform rock and the transformation-is-risky hard place?  

I’ve learned that the key to reliable and scalable transformation is an infrastructure that is fit-for-purpose – designed to meet each current and future business objective.  

Fit-for-purpose approach to modernizing core banking systems

Many banks build a target state environment assuming that all the components of their core will be fully modernized to cloud-native applications. This approach breaks down when the inevitable decision gets made that some of the applications will not be fully modernized; maybe the deadline to close a data center is approaching, maybe the budget couldn’t be secured for a rewrite, or maybe a regulator is requiring progress in a short time period.

Whatever the reason, the bank now has apps that are not horizontally scalable, not built with zero trust, and not optimized for cloud services. The apps are living in a target state that wasn’t intended for them, which leads to high cloud bills, security vulnerabilities, and manageability challenges.

That’s why a fit-for-purpose approach to modernization starts with the understanding that numerous levels of transformation are best-practice for modernizing a core banking system. Leaders will want to determine the levels of transformation they’re going to perform before they’re able to establish which platforms are fit-for-purpose. For example, you may find that you have a set of apps that need to be re-developed and another set that will be lifted and shifted to the cloud.

Once you understand the different levels of transformation needed to modernize your core banking system, you can determine which platforms need to be built.

Five fit-for-purpose platforms for banks

Let me further explain using a real-life use case. I recently worked with a customer whose core applications sat on on-premises across distributed and mainframe environments. They knew they needed to modernize all their applications, but given prioritization, wanted to keep transformation costs as low as possible.

Ultimately, they moved their applications to the cloud because of its resiliency, latency, and integration benefits. However, not every application could be completely rewritten, so they built four different patterns to accommodate the varied modernization journeys. The transformation levels and strategies for the four distinct fit-for-purpose platforms included:

• Migration platform. Some of the mainframe applications were near the end of their lifecycle and set to be replaced with a SaaS solution, implying a very low ROI on any extensive level of transformation. So, they chose to move those applications to zCloud without transformation.
• Mainframe platform. Another application on the mainframe needed to run alongside distributed applications for performance reasons. The customer chose to transform this application to Java without updating the business logic. They moved to a distributed cloud infrastructure with the transformed application on the cloud and the environment managed centrally from the cloud provider.
• Cloud-native platform. There were several applications that the customer relied on for competitive differentiation but were running on legacy technology that was difficult to adapt and change with customer preferences. For these applications, they decided to rewrite the application and update both the tech stack and the business logic. They opted to create an entirely new cloud environment with guardrails and DevOps/DevSecOps tooling to make software development effective and secure.
• Managed apps platform. Finally, the customer made use of an ERP application deployed on their mainframe.  The vendor solution was also available on the cloud with expanded functionality, so they decided to deploy the cloud version on a managed service platform.

It’s important to note that most core banking systems have at least a few portions that either have already been or could easily be containerized. This represents a fifth fit-for-purpose platform that I call a centralized container platform.

As an example, I often see Java workloads that require substantial parallel processing on separate app servers. While containerizing provides more efficient resource utilization and better scaling, it does not make the application ready for a cloud native account. I suggest creating a centralized container platform as a sort of intermediary for these apps.

By understanding the application portfolio and creating fit-for-purpose platforms for each type of workload, this customer, and others like them, can support each application with a secure, cost-effective, and performant environment.

Four ways to avoid core system transformation pitfalls

As you prepare to modernize your core systems on a fit-for-purpose infrastructure, our advice is to look at your entire portfolio of applications and the intended levels of transformation.

By comparing the required transformation levels with specific business needs and priorities, you can determine which platform(s) are appropriate for your modernization. Typically, patterns emerge. This allows you to identify the specific number and types of platforms you need to build.

As you transform your applications, reference these four tips to ensure you’re gaining momentum for transformation while avoiding the common pitfalls:

1. Steady communication. You know you’re doing something right once you’ve changed how you communicate with your app teams. For example, they should be inviting infrastructure product owners to their team-wide “big room planning” sessions to share what features are going to be available on the fit-for-purpose platforms.
2. Speed of transformation. The speed at which your application teams can adapt to the new infrastructure is a sign of whether you’re meeting your goals. The uplift to adapt should go down because there's a uniformity to how the applications within a platform function. Are your application teams able to build development environments faster than before?
3. Maintenance. As these cloud platforms mature, your overall upkeep of infrastructure should be trending downwards, so that the size of the infrastructure team should be able decrease. This should happen because of the efficient leveraging of cloud services in each platform.
4. Collaboration. Since platforms are often being built on the cloud, building the new infrastructure presents a terrific opportunity to break down the traditional compute, storage, network, and security silos to create an Agile Release Train with cross-functional, self-managed teams.

So, if you are ready to escape the need-to-transform rock and the transformation-is-risky hard place, a fit-for-purpose approach can alleviate many risks associated with modernization and help drive a reliable, scalable transformation journey.

Jonathan Cavell is the Practice Leader for the Apps, Data, and AI Practice in Kyndryl Consult.

1  Driving customer experience through IT modernization, American Banker, May 2023
2  Flipping the Odds of Digital Transformation Success, Patrick Forth, Tom Reichert, Romain de Laubier, and Saibal Chakraborty, October 2020