AI-generated video
Vibe Coding, code optimisation
and quality gates.
AI now writes large parts of the code — faster than teams can review it. This series covers what comes next: how AI-generated code can be measured, checked and secured, which metrics actually hold up, and at which point a gate is worth more than a review.
Test Coverage for AI Generated Code: Why the Percentage Is Misleading
AI writes tests fast, and derives its expectations from the very code it is supposed to check. Here is how to avoid that trap.
Quality Gates in the CI/CD Pipeline: Building Four Stages
A pipeline that only builds and ships is a conveyor belt. Here is how it becomes a checkpoint chain that stays fast and still knows when to stop.
What Is a Quality Gate? Definition, Criteria, and Common Mistakes
A quality gate is a barrier with fixed criteria that a change must pass. How to build one that actually helps instead of just being annoying.
Clean Code in the AI Era: Which Principles Stay and Which Shift
When a machine writes the code, readability does not become less important, it becomes more important. Which of the old rules count now.
Modernizing Legacy Code With AI: What Works and What Goes Wrong
AI can explain, document, and test old code, that is the underrated win. The traps sit in the rewriting.
Refactoring AI-Generated Code: An Approach That Breaks Nothing
Refactoring has almost vanished from repositories, from 21 down to 3.8 percent. How to restructure generated code without carrying its bugs along.
Identifying and Fixing Code Smells: The Ten Most Common in AI Code
A code smell is not a bug, it is a signal. Which patterns generated code reliably produces, and how to resolve each one.
Static Code Analysis: What It Finds, What It Misses, and Which Tools Fit
Static analysis checks code without running it: the cheapest error brake there is. Where its limits lie and how to introduce it into an existing codebase.
Code Review for AI Code: Six Patterns the Classic Review Misses
The classic review looks for careless mistakes. AI code has a different kind: it looks correct. A review workflow built for exactly this class of error.
Measuring Code Quality: Which Metrics Matter and Which Just Reassure
Test coverage alone says almost nothing. Which metrics give a real read on maintainability, and how ISO 25010 helps make sense of it.
Technical Debt from AI Code: What the Data Shows
Copied blocks up 81 percent, genuine cleanup nearly gone: how AI assistance shifts the maintenance burden and what helps against it.
From Prototype to Production: What a Vibe-Coded Draft Still Needs
The prototype works, everyone is thrilled, and suddenly it is supposed to go live. The twelve points standing between "it works on my machine" and "it can carry real users."
All images in this overview are AI-generated illustrations.AI transparency
Never miss an insight
Subscribe to our newsletter and get the latest articles straight to your inbox.
No spam. Unsubscribe anytime.
Ready for your AI competence certificate?
Get your AI certificate – flexible, online, with a documented certificate of participation for building AI literacy (Art. 4 EU AI Act).