DUDES / CONTINUED DEVELOPMENT OF EXISTING SYSTEMS

Legacy modernization. No big bang.

Legacy modernization, for us, means improving an old system one piece at a time: understand it first, protect it with tests, then change it. A full rewrite is rarely the right first step.

WHO IT’S FOR

Is this you?

↗01

System owners

The business runs on old code, and every change is scary.

⌘02

Teams that inherited a codebase

The original authors are gone, there’s no documentation and no tests.

⇄03

Companies on ageing technology

A framework, language version or infrastructure is reaching end of life.

THE TYPICAL SITUATION

A rewrite sounds easy until you try it.

An old system contains years of business rules nobody wrote down. A full rewrite has to rediscover all of them — often in production.

Patching alone isn’t enough either if every change breaks something. You need a way to change the system safely.

HOW WE SOLVE IT

Understand, protect, change

01

Map the codebase

Agents read the code in parallel and build a map: modules, dependencies, data flows and risk areas. A senior checks the map, because a model can misread old code.

02

Tests before changes

Current behaviour is locked in with tests, so we know what changed and what didn’t.

03

Gradual replacement

Piece by piece into a new structure (the strangler fig pattern). Old and new can run side by side until the new part is proven.

04

An honest assessment

Sometimes a part is worth rewriting. We tell you where and why — not as a general principle.

HOW THE WORK MOVES

We don’t start over just because we didn’t start it.

WHAT YOU HAVE
  • Idea
  • Lovable
  • Bolt / v0
  • Existing code
GITGit repo
DUDES ✳ WHAT WE ADD
  • ✓ Architecture
  • ✓ Backend
  • ✓ Database
  • ✓ Auth
  • ✓ Integrations
  • ✓ Testing
  • ✓ Infra
  • ✓ Security
  • ✓ Monitoring
LIVEProduction
WHAT WE CAN BUILD

What modernization can mean

  • ↗ Codebase map and risk report
  • ↗ Test coverage before changes
  • ↗ Framework and language upgrades
  • ↗ Carving pieces out of a monolith
  • ↗ Renewing an old user interface
  • ↗ Moving to modern infrastructure and CI/CD
  • ↗ Documentation for the next developer
QUALITY & SECURITY

Generated isn’t done. Verified is.

AI models make mistakes, and so do people. Our process is built to catch them before your users do: human review, tests suited to the risk and validation before release.

See what “done” means ↗
DEFINITION OF DONE ↙
✓Current behaviour locked in with tests01
✓Small, reversible changes02
✓Production monitoring after changes03
✓A rollback plan for every release04
PRICING & ENGAGEMENT

How do we get started?

A good start is a bounded assessment: the state of the codebase, the biggest risks and a sensible first step. It often fits a bounded build sprint (€5,000 + VAT for the first week, fixed price).

5 000 €+ VAT · one week, fixed price
Start a build sprint↗Pricing & ways to work ↗
FAQ

Fair questions.

Rewrite or gradual modernization?+

Usually gradual. A rewrite is justified when the old solution blocks a critical change and can’t reasonably be protected with tests. We decide after the assessment, not before.

How does AI help with old code?+

Agents read a lot of code quickly: tracing dependencies, describing behaviour and drafting tests. The results are checked, because a model can misunderstand old code.

Do we need to stop the system?+

Usually not. Changes are small and released in a controlled way, and old and new can run side by side.

FURTHER READING

Read before you buy.

LET’S MAKE IT REAL ↙

Which system scares you most?

Tell us about the system and what makes it hard to change. We’ll start with an assessment.

Ask for a codebase assessment↗