Speakylex
Learn languages with TikTok videos & AI tutor
Hey Indie Hackers — I'm Eugene, solo founder of Speakylex.
Speakylex turns TikTok-style educational videos into language practice: short lesson → tasks → writing with AI feedback → personal dictionary. Free web app + grammar tests in the browser (no sign-up).
Website: https://speakylex.com
Android: https://play.google.com/store/apps/details?id=com.speakylex.android
Why: most language apps feel like drills disconnected from how people actually consume content. I wanted learning on top of short videos you'd open anyway, then practice right after.
Stack: Flutter, Angular, .NET. Bootstrapped, no ads in the product.
Focus now: small UX wins, SEO on grammar-test pages, retention / daily loop, then a small paid install test. One channel at a time (next: AlternativeTo, etc.).
Ask:
1) Would you start from a video feed or from writing / AI chat?
2) Web-first or app-first for something like this?
Happy to answer anything about the project. Honest feedback welcome.
7 Comments
The interesting part is putting actual language practice directly after the short-form content instead of separating “content consumption” and “learning.” I’d be most curious whether that transition creates enough of a habit to drive retention, because the feed itself is easy to consume without doing the harder practice.
Totally fair point — and yeah, I still need to find that out. That flow is new, so retention around the video → practice jump is exactly what I'm watching.
One thing I'm counting on: the tasks after a video are usually small, you can pick the type, and you can come back to them later — so it's not an all-or-nothing moment right after watching.
Honestly retention is the main problem I want to understand right now. Curious if you've seen products where that "watch → then practice" jump actually worked?
That’s a useful retention question. I’d be curious to see what the early usage data says about how often people actually make that jump from watching to practicing.
I don't have a solid number to share yet — the practice-after-video loop is still early, so the sample is small
Feels like there's still a lot of work ahead to get this loop right. Thanks for the thoughtful questions.