At the European Digital Finance Conference, organised by The Banking 50 on March 5 in Amsterdam, Michał Niedźwiecki, IT Director at mBank, shared a compelling, multi-year story of transforming a core banking system—without disrupting business operations.
What emerged was not just a technical case study, but a strategic blueprint for modernization in highly complex, always-on environments.
The journey began with a simple but uncomfortable question posed by leadership: Is our core banking system modern—and if not, what are we going to do about it?
At the time, mBank was running a decades-old core system built in COBOL on a mainframe. Like many legacy banking environments, it was stable—but increasingly rigid, expensive to scale, and difficult to evolve at the pace the business demanded.
The mandate was clear: modernize the core system without slowing business growth.
Contrary to conventional thinking, Niedźwiecki emphasized that the greatest risk wasn’t transformation—it was inertia.
mBank faced mounting challenges:
Beyond technical constraints, there were strategic concerns. The bank wanted greater independence, flexibility in deployment (on-premises, hybrid, or cloud), and the ability to leverage open technologies rather than remain tightly coupled to legacy infrastructure.
A full system replacement in a single cutover—often referred to as a “Big Bang”—was considered, but quickly dismissed.
For a bank serving millions of customers and operating nearly 24/7, concentrating all risk into a single migration event was simply unacceptable.
Instead, mBank chose a fundamentally different path:
Gradual, continuous migration as a process—not a one-time event.
The transformation was built on several key principles:
This approach ensured:
A critical enabler of the transformation was the adoption of application and database sharding.
Instead of a single monolithic core, the system was split into smaller, independent units (“shards”), each handling a subset of customers and data.
This shift delivered:
What began as a transitional architecture ultimately became the target state, allowing the bank to scale dynamically by adding more shards as needed.
From a customer perspective, the migration was כמעט invisible.
Behind the scenes, however, the impact was profound:
mBank even rebuilt critical components—such as its authorization system—directly on the new platform, removing legacy inefficiencies in the process.
The scale of the transformation was significant:
This rigorous testing strategy ensured that the new environment maintained functional parity and reliability throughout the transition.
Despite the technical complexity, Niedźwiecki stressed that modernization is not primarily about technology.
Instead, success depended on:
Notably, no jobs were lost during the transformation. The same engineers who understood the legacy system became key contributors to the new architecture.
Niedźwiecki concluded with four essential lessons:
Perhaps the most powerful takeaway was that the project didn’t just change the system—it changed the organization itself.
Teams evolved, capabilities expanded, and a culture of resilience and collaboration emerged.
As Niedźwiecki reflected, modernization is as much about people and mindset as it is about architecture.
