Linguify
Conversational AI language learning app. Best Start-up, VT Hacks.
[ Visit project โ ]Summary
A language-learning app that replaces rote flashcards with AI conversation practice grounded in photos of your real surroundings.
Problem you're solving
73% of learners report anxiety using a new language in real conversations, and app-based learners on tools like Duolingo skew heavily toward reading and writing over listening and speaking.
Target user
Self-taught language learners who've plateaued on an app like Duolingo and need real conversational reps, not more vocabulary drills.
Competitors
Duolingo and Babbel run on repetitive, scripted exercises with limited free-form speaking; human tutors on iTalki are effective but expensive and require scheduling.
Insight
Learners don't need more vocabulary โ they need low-stakes conversation practice tied to things they actually see and do, with instant correction instead of waiting for a class.
Solution
Users photograph their surroundings to auto-generate contextual vocabulary, then hold adaptive spoken conversations with an AI tutor that adjusts to their level and gives real-time pronunciation and comprehension feedback.
- Next.js
- TypeScript
- FastAPI
- Python
- Firebase
- OpenAI
- Retell AI
Distribution
Built and demoed at VTHacks 12; distribution so far is the hackathon demo day and Devpost listing, ahead of a planned public launch.
Adapting to user feedback
Judges and early testers flagged that authentication friction was killing the demo flow, so onboarding was simplified before the final round โ a fix that helped win Best Startup Hack.