Introducing Koshei AI: A Voice-Driven Language University from A1 to D2 Levels

Transforming Language Learning with AI 🌟

I’m excited to share my journey in creating a voice-first AI language teacher right in your browser, leveraging cutting-edge technologies like:

• Next.js
• Gemini Audio API for Speech-to-Text (STT)
• Gemini Text-to-Speech (TTS)
• Supabase for backend support

💡 Vision: A dynamic AI “language university” rather than just another chatbot. My goal is to empower language learners with personalized, structured instruction.

🔍 Current Challenges:

• Implementing lightweight browser-native lip sync for static avatar images during TTS audio playback.
• Ensuring MediaRecorder reliability on mobile Safari.
• Enhancing voice UX for guided teaching.

Demo & Repository:
Check out the demo here and my code on GitHub.

I’m eager for your technical suggestions and architecture feedback—let’s collaborate to shape the future of language learning!

👉 Share your thoughts and let’s innovate together!

Source link