United Kingdom - English Investors Newsroom Alumni Platform
Contact us
Case study

Modernising a 220M-account retail bank to cloud-native in 18 months.

A European retail bank had already spent three years trying to replace its core in waves. We reset the program around domain-by-domain modernisation, an event backbone, and a managed cutover factory that could survive real regulatory scrutiny.

220M
customer accounts migrated
18 months
from reset to final country cutover
70%
faster launch cycle for new products
0
customer-facing incidents during cutover

The situation we walked into.

The client carried 420 downstream applications on top of a 30-year-old core, with country-specific product logic hard-coded into nightly batches. Every prior migration attempt had stalled when finance, fraud, and servicing teams discovered a dependency too late.

We reframed the objective away from 'replace the core' and toward 'make product launches and resiliency materially better'. That let the bank sequence work by customer journeys, keep regulators informed with live evidence, and prove value before the final country cutover.

A simplified view of the delivery shape, the control points that mattered, and the signals the client team used to keep the program on track.

Program workstreams

The changes that made the outcome possible.

Stream 01

Core decomposition

Deposits, cards, lending, and servicing were carved into independently releasable domains with shared policy services.

Stream 02

Event fabric

Kafka-backed canonical events replaced overnight file movement so risk, finance, and channels could subscribe in real time.

Stream 03

Cutover factory

Every rehearsal executed against the same runbooks, controls, rollback triggers, and executive command-centre dashboards.

Stream 04

Run transition

Managed services and SRE teams stood up 24x7 reliability coverage before the first production migration.

Execution rhythm

How the delivery moved from pilot to scaled operation.

Months 0-3

Program reset

A single control tower, release calendar, and regulator-ready evidence pack replaced 17 local workstreams.

Months 4-8

Foundations live

Deposits and onboarding domains launched behind APIs while the legacy core still handled settlement.

Months 9-14

Dual-run scale-up

Cards, servicing, and customer messaging ran in parallel for six countries with automated reconciliation.

Months 15-18

Country cutovers

Weekend migrations completed with rehearsed rollback points, war-room telemetry, and board-level dashboards.

Twelve months later

What changed after the transformation settled into the run.

Outcome 01

Product change moved from quarters to weeks

The retail product team can now ship rate, fee, and bundle changes through a governed API release path instead of core code freezes.

Outcome 02

Run economics improved immediately

Legacy mainframe demand dropped, incident triage was automated, and the bank closed four expensive manual reconciliation towers.

Outcome 03

Reliability became measurable

SLOs, golden signals, and resilience drills gave the COO the first real-time view of customer-impacting degradation across all markets.

"
The difference was not just better engineering. It was a delivery model that let compliance, operations, and product teams trust the move in real time.
Group CIO - European retail bank

Want the full delivery playbook for a similar transformation?

Request the deeper dive