ManyMinds React Native & Laravel Backend

📖 Project Overview

ManyMinds is a real‑time AI‑powered chat application built with a React Native front‑end and a Laravel back‑end. It showcases:

• AI chat powered by OpenAI's GPT‑4o‑realtime model.
• Live voice streaming using WebRTC (React Native WebRTC) for two‑way audio communication.
• Authentication via Laravel Sanctum.
• Rich conversation management (create, list, view, and stream messages).
• AI profile handling and prompt building services.
• Realtime session creation for voice calls.

The project is a learning sandbox for modern mobile development, real‑time communication, and server‑side AI integration.

---

✨ Key Features

Feature	Description
Chat Service	Handles sending/receiving messages, creating conversations, and streaming AI responses.
Realtime Voice Service	Sets up a WebRTC peer connection, captures microphone audio, streams voice from OpenAI, and provides a data channel for control messages.
Auth Service	Register, login, and fetch the current user using Laravel Sanctum tokens.
AI Profile Management	CRUD endpoints for AI profiles (system prompts, temperature, etc.).
Prompt Builder	Server‑side service that builds prompts for the AI model.
Realtime Session API	Generates an ephemeral token for the OpenAI realtime endpoint.
Laravel API	RESTful routes for auth, AI profiles, conversations, messages, feedback, and realtime sessions.
React Native UI	Modern mobile UI with dark mode support, gradient backgrounds, and smooth animations.

---

🛠️ Technology Stack

• Frontend: React Native (TypeScript), react-native-webrtc, axios, react-navigation
• Backend: Laravel 12 (PHP 8.2), Sanctum, OpenAI PHP SDK (openai-php/laravel)
• Realtime: OpenAI Realtime API (gpt-4o-realtime-preview-2024-12-17)
• Database: SQLite (default) – can be swapped for MySQL/PostgreSQL.
• Package Management: npm (frontend), Composer (backend)
• Development Tools: Vite (frontend bundler), Laravel Sail (Docker dev environment optional)

---

🚀 Getting Started

Prerequisites

• Node.js (>=18) and npm
• PHP (>=8.2) and Composer
• Android SDK (for Android emulator/device) – see React Native docs.
• OpenAI API key with access to the Realtime model.

Backend Setup

1. Clone the repository and navigate to the project root.
2. Copy the example env file and set your variables:

   cp backend/.env.example backend/.env

3. Edit backend/.env and add:

   OPENAI_API_KEY=your_openai_key_here

4. Install dependencies and run migrations:

   cd backend
   composer install
   php artisan key:generate
   php artisan migrate --force

5. (Optional) Start the Laravel development server:

   php artisan serve

   The API will be available at http://127.0.0.1:8000/api.

Frontend Setup

1. Install npm packages:

   cd my-app
   npm install

2. Create a .env file in my-app (copy from .env.example if present) and set the backend URL if you are not using the default localhost:

   VITE_API_BASE_URL=http://127.0.0.1:8000/api

3. Run the app on Android (or iOS):

   npm run android   # or `npm run ios`

   The Metro bundler will start and the app will launch on the emulator/device.

---

📡 API Overview

Method	Endpoint	Description
POST	/api/auth/register	Register a new user.
POST	/api/auth/login	Login and receive a Sanctum token.
GET	/api/me	Get the authenticated user.
GET/POST	/api/ai-profiles	Manage AI profiles.
GET	/api/conversations	List conversations.
POST	/api/conversations	Create a new conversation.
GET	/api/conversations/{id}	View a conversation.
GET	/api/conversations/{id}/messages	List messages for a conversation.
POST	/api/conversations/{id}/messages	Send a message (handled by ChatController).
POST	/api/conversations/{id}/realtime-session	Generate an ephemeral token for the OpenAI Realtime endpoint.
POST	/api/messages/{id}/feedback	Submit feedback for a message.

---

🎤 Realtime Voice Flow

1. The front‑end calls chatService.getRealtimeSession(conversationId) to obtain an ephemeral token.
2. RealtimeVoiceService creates a RTCPeerConnection and captures the microphone via mediaDevices.getUserMedia.
3. An SDP offer is sent to OpenAI's Realtime endpoint (https://api.openai.com/v1/realtime).
4. The response SDP answer is applied, establishing a two‑way audio stream.
5. A data channel (oai-events) is used for control messages (e.g., sending text prompts during a call).

---

🧪 Testing & Verification

• Backend: Run php artisan test to execute the Laravel test suite.
• Frontend: Use the React Native testing library or manually test the UI on an emulator/device.
• Realtime: Verify voice streaming by starting a call from the chat screen; you should hear the AI's spoken response and see the microphone indicator.

---

🤝 Contributing

1. Fork the repository.
2. Create a feature branch (git checkout -b feature/awesome-feature).
3. Make your changes and ensure both backend and frontend tests pass.
4. Open a Pull Request describing the changes.

---

📄 License

This project is licensed under the MIT License – see the LICENSE file for details.

---

📞 Contact

For questions or suggestions, feel free to open an issue or contact the repository maintainer.