Articles about how I work
Short essays on how I connect QA, requirements, delivery, and complementary AI practice in real work.
Why quality assurance and requirements matter more in the age of AI agents
AI agents shift the bottleneck in software projects: code can be produced faster, but domain clarity, quality, and control become even more important.
Metaprompting: The prompt that writes better prompts
Metaprompting means not asking an AI for the final answer right away, but first improving the task itself. The result is clearer prompts and better outputs.
Reducing token costs in AI agents without losing quality
Token optimization is not only a prompt-writing problem. In production agents, the biggest leverage comes from architecture: compact prompts, context management, RAG, caching, and model routing.
Three Claude Code skills that are changing my QA work
Writing test concepts, running E2E tests, and finding coverage gaps – three specialized skills that automate recurring QA tasks and make domain expertise reproducible.
How I systematically improve software quality – my approach as an IT consultant
Quality problems rarely appear suddenly – they develop gradually. My structured five-step approach makes quality measurable, identifies risks early and guides projects to a stable go-live.
What I have actually learned after several years in QA
Quality does not emerge during testing alone. It emerges in the system, in the process and in the way a team thinks.
How I built a SaaS without a traditional developer background - using AI and a QA mindset
NutriKompass started with a concrete operational problem in a therapeutic care setting and the question of how far AI, product thinking and QA experience can take you today.