Personal insight
How I use AI to actually implement ideas
A practical workflow from a concrete problem to a result that survives real use.
The process starts with a problem, not a prompt
I first describe the situation, the user and what should become easier. AI becomes useful when the context is concrete.
My practical workflow
I clarify the goal, provide relevant context, build a small first version, review the output and test it in practice.
Context determines quality
Constraints, examples and a clear definition of success produce far better results than a clever one-line prompt.
A first version is not a finished result
AI accelerates implementation, but judgement, testing and iteration remain essential.
A concrete example: Dailumo
I missed having a small month view right in my Mac menu bar. With support from AI, I built Dailumo and started using the first version myself. Everyday use revealed the details that needed attention: the alignment of the day highlight, month and year selection, and an easy-to-reach launch-at-login setting. Those observations are what turn a first version into a useful tool.
Learning through structured development
My AI programme from July 2025 to July 2026 deepens the practical perspective and helps me connect tools with real organisational needs.
Explore the projects behind these lessons
See FlowTime, LeasingTracker and the products currently in development.