A voice-driven English speaking practice app that helps non-native speakers improve their conversational English skills through AI-powered coaching.
PronouncePal is a fully voice-based English learning app where you:
- Talk naturally - Just click the mic and speak
- Get instant feedback - AI coach responds with corrections and encouragement
- Practice conversations - Have real dialogues, not scripted exercises
- Hear native pronunciation - AI speaks back with natural voice
The AI coach remembers your conversation and provides gentle grammar corrections while keeping things fun and engaging.
- Frontend: React + Vite + Tailwind CSS
- Backend: Node.js + Express
- AI: Google Gemini 2.5 Flash
- Voice: Web Speech API + ElevenLabs TTS
- Deployment: Vercel
git clone https://github.com/Mujaheed56/pal.git
cd palnpm install
cd backend
npm install
cd ..Create .env in root:
VITE_API_URL=http://localhost:5000/apiCreate backend/.env:
PORT=5000
GEMINI_API_KEY=your_gemini_api_key_here
ELEVENLABS_API_KEY=your_elevenlabs_api_key_hereTerminal 1 - Frontend: Terminal 1 - Frontend:
npm run devTerminal 2 - Backend:
cd backend
node server.jsOpen http://localhost:3000 and start speaking!
- Gemini API: https://aistudio.google.com/app/apikey
- ElevenLabs: https://elevenlabs.io (free tier available)
- Click the microphone button
- Allow mic permissions
- Speak naturally in English
- Listen to AI coach's response
- Keep the conversation going!
✅ Real-time speech recognition
✅ Natural AI conversations
✅ Grammar corrections
✅ Text-to-speech responses
✅ Conversation memory
✅ Mobile responsive
Built for hackathon 🚀