Skip to content

Legacy modernisation - case study - 2023

quantiq

Modernising a clinical analytics platform without a freeze

Client
quantiq · clinical analytics
Team
7 engineers, 1 QA, 1 product owner
Engagement
Time & materials · embedded
Compliance
GDPR · ISO 27001

The numbers after launch.

Release cadence, from quarterly
WeeklyRelease cadence, from quarterly
Planned downtime
0 minPlanned downtime
Hosting cost
−62%Hosting cost
Full migration
9 monthsFull migration

01 / Brief

quantiq's analytics platform served hospital trusts from a single Java monolith written over twelve years. Every release needed a weekend window and a full regression cycle, so changes shipped once a quarter. The product team had a backlog of requests from customers it could not deliver, and a full rewrite was ruled out: the platform could not stop while one was built.

02 / Solution

We migrated the platform piece by piece behind a routing layer, starting with the reporting module that changed most often. Each extracted service ran alongside its monolith equivalent, compared outputs on live traffic, and took over only once the results matched. Deployment moved to containers with automated tests on every change, which is what made weekly releases safe. The monolith was switched off nine months later without a single planned outage.

Technology used: Java, Kotlin, Python, FastAPI, PostgreSQL, Kubernetes

— Tech stack

Built with what holds up at scale.

  • Java
  • Kotlin
  • Python
  • FastAPI
  • PostgreSQL
  • Kubernetes

— Client

We kept shipping to hospitals the whole time. Nobody outside the team noticed the platform underneath had changed.
— CTO, quantiq

More customer cases

More work, same standard.

All customer cases
Fintech

Platform engineering  •  2025

Same-day reconciliation for a payments provider

A nightly batch job and three spreadsheets replaced by an event-driven ledger that reconciles card settlements within the hour.

helixpay