Too important to switch off. Too old to keep building on.
Almost every organisation has a system like this. It has run for years, it keeps the business going, and nobody dares change anything in it. The original developer is gone, the documentation is incomplete, and every change feels like surgery without an X-ray.
Why a rewrite is usually the wrong answer
An old system looks bad because you can see that it has been working. Its special cases, exceptions and odd branches hold twenty years of knowledge about your business — knowledge that is written down nowhere else.
A full rebuild throws that knowledge away and has to rediscover it at great expense. Usually while the business keeps running, usually under time pressure, usually with people who never saw the original.
That is why we prefer to modernise step by step.
Typical starting points
- The original developer or vendor is no longer available
- The technology no longer receives security updates
- New requirements cannot be built without putting the existing system at risk
- The system depends on one person who will retire in the foreseeable future
- Compliance requirements (accessibility, GDPR, NIS2, DORA) can no longer be met
Getting started: Technical Review
Two days, fixed price. We look at code, architecture, dependencies and the runtime environment and deliver a written assessment: condition, risks, options for action with effort estimates. The document is yours — including if you carry on with someone else.
How we go about it
- 1
Inventory
What does the system really do? Which parts are actually used? (Often a third fewer than assumed.)
- 2
Risk map
Where is it dangerous, where is it merely awkward, where does it not matter?
- 3
Safety net
Characterisation tests for critical behaviour, before anything is touched.
- 4
Incremental replacement
Functional areas are carved out and replaced one at a time while the old system keeps running. Every step can be rolled back on its own.
- 5
Switch-off
The old system only goes offline once nothing accesses it any more.