The app no one will ever use (except my daughter)
Last week, my daughter came home with a vocabulary list. Fifty English words. Test on Friday.
Last week, my daughter came home with a vocabulary list. Fifty English words. Test on Friday.
You know the drill. Parent sits down with the list. Kid stares at the ceiling. "What's the English word for 'hond'?" Silence. "Dog." "Oh yeah, dog." Repeat forty-nine more times. Everyone's bored. Nobody's learning.
I've done this routine dozens of times. And every time I think: there has to be a better way.
This time, I decided to build one.
The idea
What if I made a simple Duolingo-style app? Multiple choice questions, typed answers, instant feedback. Gamified enough to not feel like homework.
But the real magic trick: what if I could just photograph the vocabulary list and have AI extract all the words?
No manual typing. No copying fifty words into a spreadsheet. Just snap a picture of the school worksheet and go.
The build
I fired up Laravel with React and started vibing.
Three hours later, I had:
- A quiz system with multiple choice and typed answers
- Score tracking and progress bars
- Sound effects for right and wrong answers
- A hint system (reveal letters one by one)
Instant feedback keeps it engaging — green for correct, move on. Wrong answer? Show the correct one and keep going.
The next morning, I added the scan feature. Upload a photo of the vocabulary list, Claude extracts the Dutch-English word pairs, and populates the database. It even generates "confusables" — similar-sounding words to use as wrong answers in multiple choice.
Upload a photo of the word list... and AI extracts all the words automatically.
Total build time: maybe six hours spread across two evenings.
The result
My daughter opened the app, typed her name, and started quizzing herself.
No parent required.
She went through the entire list three times that evening. Not because I asked her to — because the instant feedback loop made it feel like a game, not a chore. Green flash for correct, encouraging sound, next question.
Friday came. She aced the test.
The point
This isn't a story about building the next Duolingo. It's not even a particularly impressive app — it's a simple quiz with a neat AI trick.
But here's what strikes me: a year ago, I wouldn't have built this.
Not because I couldn't. Because the effort-to-payoff ratio felt wrong. A custom app for one vocabulary test? For one kid? That's a weekend project for a Tuesday evening problem.
There's been a lot of discussion lately about whether AI is replacing developers or making us obsolete. I think that misses the point entirely.
Vibe coding changed that math.
With the right AI tools, the friction dropped low enough that building a personal tool became the obvious solution. I didn't plan architecture. I didn't write specs. I described what I wanted, iterated fast, and shipped something useful in hours.
That's the real unlock. Not that we can build bigger things — but that we can build small things for small problems. Personal tools that scratch personal itches.
My daughter doesn't know she's using a Laravel + React app. She just knows that practicing vocabulary is less boring now.
And honestly? That's the best product feedback I've ever received.
What small problem would you solve if building was effortless? Sometimes the best apps are the ones nobody else will ever use.
More on vibe coding: I'm documenting my journey building Dotmap and Expensio — two larger projects where the same "describe and iterate" approach scales up. And I wrote about a CLI tool that summarizes 60-minute tech talks in 34 seconds.
Daniel Plomp
Daniel Plomp
I offer hands-on consulting to help you resolve technical challenges and improve your CMS implementations.
Get in touch if you'd like support diagnosing or upgrading your setup with confidence.
