Audex - AI-Powered Audio Transcription

A full-stack web application that converts audio files and voice recordings to text using AI-powered transcription (Whisper Large V3 model via Groq API).

🌐 Live Demo

• Frontend: audexai.vercel.app
• Backend: audex-backend.onrender.com

---

📖 Project Overview

Audex is a full-stack web application that converts audio files and voice recordings into text using AI-powered transcription. It leverages the Whisper Large V3 model via Groq's ultra-fast inference API to deliver lightning-fast, highly accurate transcriptions in 28+ languages.

---

🔧 How It Works

Architecture Flow

┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│   User Upload   │     │  Spring Boot    │     │    Groq API     │
│   Audio File    │────▶│    Backend      │────▶│  (Whisper V3)   │
│   or Record     │     │                 │     │                 │
└─────────────────┘     └─────────────────┘     └─────────────────┘
                                │                       │
                                │◀──────────────────────┘
                                │    Transcription JSON
                                ▼
                        ┌─────────────────┐
                        │  React Frontend │
                        │  Display Result │
                        └─────────────────┘

Step-by-Step Process

1. Audio Input: User uploads an audio file (MP3, WAV, M4A, FLAC, OGG, WebM) OR records live audio using browser microphone
2. Language Selection: User selects transcription language from 28 supported languages
3. API Request: Frontend sends audio as multipart/form-data to Spring Boot backend
4. Groq Processing: Backend forwards the audio to Groq's Whisper API for transcription
5. Response: Transcribed text is returned and displayed with word/character stats
6. Export: User can copy text or download as .txt, .doc, or .json

---

✨ Key Features

Feature	Description
🎤 Live Recording	Record audio directly in browser using MediaRecorder API
📁 File Upload	Drag & drop or click to upload audio files
🌍 28+ Languages	English, Spanish, French, Hindi, Chinese, Arabic, etc.
🌙 Dark/Light Theme	Toggle between themes with smooth transitions
📊 Statistics	Word count, character count, language info
📋 Multiple Export	Copy to clipboard, download as TXT/DOC/JSON
🎨 Waveform Visualizer	Real-time audio visualization during recording
⚡ Fast Processing	Groq's inference is 10x faster than OpenAI

---

🛠️ Tech Stack

Frontend

Technology	Purpose
React 19	UI library with hooks
Vite 6	Fast build tool & dev server
Axios	HTTP client for API calls
FileSaver.js	Client-side file downloads
CSS Variables	Dark/Light theme system
Web Audio API	Live recording & waveform visualization

Backend

Technology	Purpose
Spring Boot 4.0	Java REST API framework
Java 21	Latest LTS version
Spring WebFlux	Reactive HTTP client
RestTemplate	Multipart file forwarding
Groq API	AI transcription (Whisper Large V3)

Deployment

Service	Component
Vercel	Frontend hosting (React)
Render	Backend hosting (Docker/Java)
GitHub	Source control & CI/CD trigger

---

🚀 Novelty & Unique Aspects

1. Groq-Powered Speed

Unlike traditional OpenAI Whisper API, Audex uses Groq's LPU (Language Processing Unit) which provides:

• ~10x faster inference than GPU-based solutions
• Near real-time transcription
• Free tier with generous limits

2. Full-Stack Java + React Architecture

• Modern Spring Boot 4.0 with Java 21
• Clean separation of concerns
• Production-ready Docker deployment

3. Browser-Native Recording

• Uses MediaRecorder API for recording
• Web Audio API for real-time waveform visualization
• No external dependencies for audio capture

4. Multi-Format Export

• Plain text (.txt)
• Microsoft Word compatible (.doc)
• Structured JSON with metadata (.json)

5. Privacy-First Design

• Audio files are processed and immediately deleted
• No permanent storage on server
• Secure HTTPS communication

---

📁 Project Structure

Audex/
├── Audex-Frontend/              # React + Vite
│   ├── src/
│   │   ├── App.jsx              # Main component
│   │   ├── AudioUploder.jsx     # Core transcription UI (700+ lines)
│   │   ├── App.css              # Theming & styles (1500+ lines)
│   │   └── index.css            # Global styles
│   ├── package.json
│   └── vite.config.js
│
├── Audex-Backend/               # Spring Boot
│   ├── src/main/java/com/audio/transcribe/
│   │   ├── AudioTranscribeApplication.java
│   │   ├── TranscriptionController.java   # POST /api/transcribe
│   │   ├── WebConfig.java                 # CORS configuration
│   │   └── WebClientConfig.java
│   ├── Dockerfile
│   └── pom.xml
│
└── README.md

---

🚀 Getting Started

Prerequisites

• Node.js 18+
• Java 21+
• Maven 3.9+

Backend Setup

cd Audex-Backend
./mvnw spring-boot:run

The backend will start on http://localhost:8080

Frontend Setup

cd Audex-Frontend
npm install
npm run dev

The frontend will start on http://localhost:5173

---

🔌 API Endpoint

POST /api/transcribe

Request:

Content-Type: multipart/form-data
- file: (audio file)
- language: "en" | "es" | "hi" | ... (optional, default: "en")

Response:

{
  "text": "Transcribed audio content..."
}

---

🌍 Supported Languages

Language	Code	Language	Code
English	en	Japanese	ja
Spanish	es	Korean	ko
French	fr	Chinese	zh
German	de	Arabic	ar
Italian	it	Hindi	hi
Portuguese	pt	Turkish	tr
Dutch	nl	Vietnamese	vi
Polish	pl	Thai	th
Russian	ru	Indonesian	id
Swedish	sv	Ukrainian	uk
Danish	da	Czech	cs
Finnish	fi	Greek	el
Norwegian	no	Hebrew	he
Romanian	ro	Hungarian	hu

---

🎯 Use Cases

1. Students - Transcribe lectures and study materials
2. Journalists - Convert interviews to text
3. Content Creators - Generate subtitles and captions
4. Accessibility - Make audio content accessible
5. Researchers - Transcribe qualitative interviews
6. Podcasters - Create show notes and transcripts

---

🏗️ System Architecture

┌────────────────────────────────────────────────────────────┐
│                         FRONTEND                           │
│  ┌──────────────────────────────────────────────────────┐  │
│  │  React Components                                     │  │
│  │  • Theme Toggle (Dark/Light)                         │  │
│  │  • File Upload with Drag & Drop                      │  │
│  │  • Live Recording with Waveform                      │  │
│  │  • Language Selector (28 options)                    │  │
│  │  • Result Display with Stats                         │  │
│  │  • Export Options (Copy/Download)                    │  │
│  └──────────────────────────────────────────────────────┘  │
│                           │                                 │
│                     Axios POST                              │
│                           ▼                                 │
└────────────────────────────────────────────────────────────┘
                            │
                      HTTPS Request
                            │
┌────────────────────────────────────────────────────────────┐
│                         BACKEND                             │
│  ┌──────────────────────────────────────────────────────┐  │
│  │  Spring Boot REST Controller                          │  │
│  │  • Receive multipart file                            │  │
│  │  • Create temp file                                  │  │
│  │  • Forward to Groq API                               │  │
│  │  • Return transcription                              │  │
│  │  • Delete temp file                                  │  │
│  └──────────────────────────────────────────────────────┘  │
│                           │                                 │
│                     RestTemplate                            │
│                           ▼                                 │
└────────────────────────────────────────────────────────────┘
                            │
                      HTTPS Request
                            │
┌────────────────────────────────────────────────────────────┐
│                      GROQ API                               │
│  ┌──────────────────────────────────────────────────────┐  │
│  │  Whisper Large V3 Model                               │  │
│  │  • 10x faster than GPU inference                     │  │
│  │  • High accuracy transcription                       │  │
│  │  • Multi-language support                            │  │
│  └──────────────────────────────────────────────────────┘  │
└────────────────────────────────────────────────────────────┘

---

🤝 Contributing

Contributions, issues, and feature requests are welcome!

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

---

📝 License

This project is open source and available under the MIT License.

---

👨‍💻 Author

Ayush Gupta

• GitHub: @ayuxsh009

---

⭐ Star this repository if you found it helpful!