The idea
I've been a Duolingo user for years but kept thinking — why does this only work for languages? The gamification loop is genuinely effective: short lessons, hearts, streaks, instant feedback. But if you want to learn coffee brewing chemistry, aviation basics, or niche professional terminology, there's nothing like it.
So I built Learn Dino. You type any topic, AI generates a complete structured course in 30 seconds — explainer content before each unit, then a gamified quiz loop. Like Duolingo but for literally anything.
It's live on iOS (https://apps.apple.com/app/id6763704798), Android (https://play.google.com/store/apps/details?id=com.learndino.app) and web (https://www.learndino.com).
What was harder than expected
Course generation consistency. Getting the AI to produce reliable structured output across any topic — from coffee brewing to regulatory compliance — required significant prompt engineering. Edge cases like very niche topics or ambiguous one-word prompts, all needed handling.
Answer scrambling. Embarrassingly simple bug — the correct answer was always in the same position. Early users noticed immediately. Fisher-Yates shuffle, five lines of code, fixed in an hour. But it cost me trust with early users before I caught it.
Understanding the needs of the users: I spent weeks adding things nobody uses now, because I convinced myself they were necessary. A few honest conversations with real feedback would have saved weeks. That's hard to get from friends frankly, everybody says it's great but they don't come back.
Where I am now
Small but real user base. Watching D7 retention as the north star metric before spending seriously on paid acquisition. First video ad creatives in production. Target is 1,000 active users then investment to bring in proper product and design expertise.
Happy to answer any questions. Brutal feedback welcome.
Top comments (0)