🎙️ Voice AI Assistant

A real-time voice conversational assistant powered by OpenAI GPT-4 and Text-to-Speech.

✨ Features

• 🗣️ Voice Input - Browser-based speech recognition (Spanish)
• 🤖 AI Responses - GPT-4 powered conversational AI
• 🔊 Voice Output - OpenAI TTS with 6 voice options
• 💾 Persistent History - SQLite database for messages and users
• ⚡ Real-time Streaming - Streamed AI responses

🛠️ Tech Stack

Layer	Technologies
Frontend	React 19, TypeScript, Vite, Material-UI
Backend	Express.js, TypeORM, SQLite
AI	OpenAI GPT-4, OpenAI TTS
Validation	Zod

📁 Project Structure

src/
├── server/                    # Backend
│   ├── core/errors/           # Custom error classes
│   ├── application/dtos/      # Request validation (Zod)
│   ├── infrastructure/        # Database entities
│   ├── presentation/          # Middlewares & responses
│   ├── services/              # Business logic
│   └── controllers/           # Route handlers
│
├── features/                  # Frontend (feature-based)
│   ├── auth/                  # Authentication context
│   ├── chat/                  # Chat logic & components
│   └── voice/                 # TTS & speech recognition
│
└── shared/                    # Shared utilities
    ├── api/                   # Centralized API client
    └── components/            # ErrorBoundary, etc.

🚀 Getting Started

Prerequisites

• Node.js 18+
• OpenAI API Key

Installation

git clone <repository-url>
cd voice-ai-app

# Frontend dependencies
npm install

# Backend dependencies
cd src/server && npm install

Configuration

Create .env in src/server/:

OPENAI_API_KEY=sk-your-api-key
PORT=3001
NODE_ENV=development

Running

# Terminal 1 - Backend
cd src/server && npm run dev

# Terminal 2 - Frontend
npm run dev

Open http://localhost:5173

📡 API Endpoints

Method	Endpoint	Description
POST	/api/users	Create/update user
POST	/api/openai	Send message, get AI response + audio
GET	/api/messages/:userId	Get user's message history
GET	/api/voices	List available TTS voices
GET	/api/health	Health check

Request Example

curl -X POST http://localhost:3001/api/openai \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello", "voiceId": "echo", "userId": "user-1"}'

Response Format

{
  "success": true,
  "data": {
    "text": "AI response text",
    "audio": "base64-encoded-audio",
    "messageId": "uuid"
  },
  "timestamp": "2024-01-01T00:00:00.000Z"
}

🏗️ Architecture

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│   Browser   │────▶│   Express   │────▶│   OpenAI    │
│   (React)   │◀────│   Server    │◀────│   API       │
└─────────────┘     └──────┬──────┘     └─────────────┘
                           │
                    ┌──────▼──────┐
                    │   SQLite    │
                    │  (TypeORM)  │
                    └─────────────┘

🔒 Security

• ✅ Input validation with Zod schemas
• ✅ Rate limiting (100 req/15min, 20 req/min for chat)
• ✅ Centralized error handling
• ✅ Request logging

🎤 Available Voices

ID	Name
alloy	Alloy
echo	Echo
fable	Fable
onyx	Onyx
nova	Nova
shimmer	Shimmer

📄 License

MIT