DualWave — AI-powered real-time communication [Under Construction]

An AI-powered real-time communication app that lets people communicate across languages using text, voice, video calls, and face-to-face modes — preserving the speaker's natural voice via AI voice cloning.

Core features

• Real-time multilingual voice and text communication
• Low-latency streaming pipeline for live calls
• AI voice cloning for natural translated speech
• Support for text chat, voice, video, and face-to-face modes

Core stack

• Mobile: React Native CLI + TypeScript
• Styling: NativeWind / Tailwind
• State: Zustand
• Navigation: React Navigation
• Backend: Node.js + Express
• Real-time: Socket.io / WebSockets
• Video: WebRTC
• Database: PostgreSQL + Prisma
• Background jobs: Redis + BullMQ
• Storage: AWS S3 / Cloudflare R2
• Translation: DeepL / LLMs
• STT: Whisper
• TTS / Voice cloning: ElevenLabs
• Business model: Freemium / Pro / Enterprise

Important architecture — AI communication pipeline

User speaks → Audio capture → Whisper (STT) → Text → Translation (LLM / DeepL) → Translated text → ElevenLabs TTS (cloned voice) → Translated audio → Recipient

Notes

• For live calls the pipeline must be streaming/low-latency (not per-utterance batch processing).
• Voice cloning must respect user consent, privacy, and legal constraints.

Quick start (project local dev)

Requirements: Node 14+, React Native environment.

Install and run:

npm install
npm start
npm run android   # or: npm run ios (macOS)

Run tests:

npm test

Project layout (high level)

• src/ — app source (components, screens, navigation, services)
• android/, ios/ — native projects
• __tests__/ — Jest tests

Contributing

• Open a focused pull request with a clear description of changes.

License: MIT