1
1 Comment

I built a language learning app from movie clips, won $30K at Hungary's biggest startup competition, and now have 300K+ users

It started because I kept pausing Netflix to look up words and forgetting them five minutes later.

So I built an iOS app for myself. Tap any word in a subtitle while streaming, get an instant translation, save it, review it later with spaced repetition. Simple.

Then I realized streaming subtitles alone weren't enough.

I started curating short movie clips as standalone lessons: watch a 30-second scene, quiz yourself on the vocabulary, repeat. That's when it clicked. Passive watching became active learning.

From there it kept growing. What was a personal subtitle tool turned into a full language learning app: think Duolingo, but built around real movie clips instead of cartoon exercises. Curated lessons, quizzes, spaced repetition flashcards, pronunciation practice, all from real native content.

Along the way I entered Hungary's biggest business competition and won the $30K grand prize.

That gave me the push to go all in: Android, Chrome extension, 20+ languages.

Where it's at now:

  • iOS, Android, and Chrome
  • 20+ languages
  • Curated movie clip lessons with quizzes
  • Spaced repetition so words actually stick
  • Interactive subtitles for streaming

The hardest part was deciding what NOT to build. Language apps try to do everything. I wanted Wordy to do one thing well: help you learn from content you actually enjoy.

Anyone else tried learning a language from movies or TV? Curious what worked for you and what didn't.

on February 28, 2026
  1. 1

    This is a great angle — movie clips give you built-in context + repetition “for free”.

    One thing I’ve noticed (I’m building in the language-learning space too): for speaking-related features, the biggest unlock is narrowing the user’s focus to one observable thing per attempt (one syllable / one stress / one sound), otherwise they just repeat the whole sentence and don’t know what changed.

    Curious: how do you help users verify they improved on the specific thing (vs. just “felt better”) after a few reps?