SyncAI is an intelligent speech-to-text transcription tool built on OpenAI's Whisper automatic speech recognition system. Featuring a user-friendly Streamlit interface, it provides accurate transcriptions with advanced features like pause detection and confidence scoring.
- Modern Streamlit UI - Clean, intuitive interface for batch audio file processing
- Advanced Analysis
- Automatic pause detection
- Word-level confidence scoring with color coding
- Average confidence score calculation
- Translation Support - Direct translation to English
- Privacy-First Design
- Fully offline processing
- No cloud uploads
- Suitable for sensitive/clinical data
| Start Screen | Results View |
 |
 |
- Python 3.7+
- pip
- Git
- PyTorch (
pip install torch torchvision torchaudio)
- Clone the repository
git clone https://github.com/whitedevil-glitch/syncai
cd syncai- Install dependencies
pip install -r requirements.txt- Launch the application
streamlit run Transcribe.py💡 Tip: Create a desktop shortcut for easy access by following these instructions
- Launch the application
- Upload one or multiple audio files using the file picker
- Select your preferred model:
- Use "large" for highest accuracy
- Use "small" or "medium" for faster processing
- Configure additional parameters as needed
- Process your files and download the transcripts
- Transcripts are also automatically saved to the local
transcriptsfolder
- Transcripts are also automatically saved to the local
| Error | Solution |
|---|---|
| Request failed with status code 403 | Run the app with the flag: --server.enableWebsocketCompression=false |
Contributions are welcome! Please feel free to submit a Pull Request.
For questions, feedback, or support:
- Email: saishankar2803@gmail.com
- GitHub Issues: Create an issue
SyncAI builds upon these excellent open-source projects:
- OpenAI Whisper - Speech recognition system
- Pyannote Audio - Audio analysis
- Whisper UI - UI components
- Whisper Streamlit - Idea Inspiration
This project is licensed under the MIT License - see the LICENSE file for details.
Built with ❤️ as a college mini-project