Modernising Heritage Financial Infrastructure: How PayPoint Accelerated Cloud Transformation with Hybrid AI

How PayPoint unblocked 25-year-old legacy systems by deploying an AI-augmented modernisation factory to rationalise a 400-workload datacentre estate, eliminate critical operational risk, and build the engineering foundations for rapid product innovation.

"This was about taking a pragmatic approach to dealing with legacy technical debt—paying it down systematically to get our business into an agile, scalable position for the future without committing to high-risk, multi-year total rewrites."

Simon Coles, Managing Director, Digital Payments & Open Banking, PayPoint

400+

Estate workloads rationalised through pragmatic triage & discovery

13x

Surge in delivery velocity (from 2 to 26 services in one month)

150+

Redundant items retired eliminating unnecessary rework

100%

Standardised observability across all refactored workloads

The Challenge

25 years of accumulated technical debt vs. business agility

As a cornerstone of UK retail payments and transaction processing, PayPoint supports critical everyday transactions across thousands of retail locations, housing associations, and utility providers.

Over decades of rapid growth, the underlying infrastructure had become constrained by an ageing physical datacentre estate:

  • Undocumented heritage systems: Hundreds of business-critical workloads were running on outdated languages (including VB6 and VB.NET) with sparse documentation, leaving vital operational knowledge trapped in the minds of a few senior specialists.
  • The "rip-and-replace" dilemma: A complete ground-up rewrite was commercially unviable and carried unacceptable downtime risk for live financial settlement pipelines. Yet, traditional manual refactoring was moving too slowly, creating delivery bottlenecks and team frustration.
  • Naive AI pitfalls: Early internal experiments attempting to automate code conversion purely with off-the-shelf AI tools fell short; archaic syntax and a lack of system context produced unverified outputs that still demanded heavy manual rework.

PayPoint required a pragmatic modernisation programme that could systematically decommission legacy dependencies within strictly controlled operating budgets, whilst establishing a modern, cloud-native engineering foundation.

The Solution

A "modernise where necessary, migrate where possible" delivery model

PayPoint and Build Circle partnered to establish a repeatable, high-throughput modernisation pipeline, shifting focus from theoretical architectural debates to high-velocity delivery.

1. Estate rationalisation & domain discovery

  • De-scoping the legacy footprint: Interrogated the raw 400-item datacentre inventory to isolate actively running systems from obsolete processes, identifying those that were inactive, duplicates, or already replaced by modern components.
  • Event-storming the transaction lifecycle: Ran collaborative domain discovery workshops with operational and accounting teams to map payment journeys independently of legacy code, ensuring newly architected services were designed around business requirements rather than old technical workarounds.

2. A pragmatic hybrid AI migration model

  • AI-assisted code deconstruction: Deployed AI tooling to parse undocumented legacy codebases and generate baseline functional prototypes (MVPs) in modern language frameworks.
  • Augmented engineering delivery: Senior engineers refined, hardened, and optimised the AI-generated baselines, collapsing rewrite timeframes from months into days per application.
  • Velocity breakthrough: Validated the methodology during a dedicated delivery sprint, surging output from just 2 completed services a month to 26 in a single August cycle.

3. Modern cloud foundations & standardised telemetry

  • Establishing the minimum quality bar: Enforced that every modernised service deployed into the continuous delivery (CI/CD) pipeline incorporated standardised OpenTelemetry logging, tracing, and health metrics.
  • Centralised observability: Replaced siloed, disparate monitoring tooling with a unified operational dashboard, giving PayPoint complete estate-wide visibility into transaction health and system performance.

Business Outcomes

Measurable risk reduction and operational momentum

  • Eliminated core "key person" risk: Transitioning legacy code into modern language stacks and automated CI/CD pipelines freed PayPoint from reliance on isolated institutional memory.
  • Predictable delivery at scale: Proved that legacy estate modernisation can be executed rapidly without compromising platform stability. 
  • Clear route off physical datacentres: Established a reliable, progressive path away from costly on-premise hardware and into scalable cloud architecture, aligning future operating costs directly with demand.
  • Data-driven engineering culture: Standardised telemetry and consistent deployment workflows enabled PayPoint’s transformation leadership to implement and track centralised DORA metrics across teams.
  • Runway for commercial innovation: Modernising core transaction processing created the reliable technical foundation needed to launch next-generation capabilities, including automated settlements, open banking, and community cash banking services.

Ready to tackle enterprise technical debt without high-risk rewrites?

Discover how Build Circle helps technical leaders combine pragmatic architecture with AI-augmented delivery to modernise critical legacy systems.

More case studies_

No items found.
View all case studies

Contact us_

Whether you’re modernising, migrating, scaling, or strengthening resilience, we’ll help you deliver faster, safer, and with stronger teams.

Let's talk about your roadmap.

Cheers, we will get back to you shortly.
Uh oh, something messed up. Try refreshing the page and submitting again.