QuastineTime

Free Qualitative tools for researcher - Fully Automated Multi-lingual Audio Transcriber and Natural Language Processing Tools using Whisper OpenAI Model

Local Audio Transcription & Text Analysis App (Standalone, No Admin Required)

QuastionTime is a local Flask-based desktop web app for audio transcription and downstream text analysis using OpenAI Whisper, NLTK, and Python. It runs entirely offline, uses a bundled FFmpeg, and can be distributed as a single Windows .exe that requires no admin privileges.

Features

• Audio transcription using OpenAI Whisper (CPU-based)
• Text preprocessing and stopword handling (NLTK)

Automatic generation of:

• Transcription text
• Summary files
• Statistics outputs

Local web interface (Flask), Fully offline after installation, No system FFmpeg installation required and Works on locked-down laptops (no admin access)

Architecture Overview

User Browser -> Local Flask Server (127.0.0.1) -> Whisper Transcription Engine -> Text Processing (NLTK) -> Local Output Files (Audio / Summary / Statistics)

All components run locally on your machine. No cloud calls are made.

Running the App (End Users)

✅ Requirements

Windows 10 or newer (64-bit) (for standalone), or pure Python console and requirements for Linux / OSX / Android

▶️ Run the Standalone EXE

• No admin rights required
• No Python installation required

Copy QuastionTime.exe to a writable folder, for example:

• Desktop
• Documents
• Downloads
• Double-click QuastionTime.exe

Open your browser and go to: http://127.0.0.1:5000

That’s it. 🎉

📁 Where Files Are Stored

All runtime data is written to a user-writable location:

%LOCALAPPDATA%\QuastionTime\

Inside that folder you will find:

Audio
Summary
Statistics
transcribe_text\

This ensures the app works even on restricted corporate or university laptops.

Developer Setup (Python Source)

1️⃣ Clone the Repository git clone https://github.com/akirawisnu/QuastionTime.git

cd QuastionTime

2️⃣ Create Virtual Environment

py -m venv .venv

..venv\Scripts\activate

3️⃣ Install Dependencies pip install -U pip setuptools wheel pip install -r Materials/requirements_file.txt

4️⃣ Run from Source python QuastionTime_standalone_ready.py

Then open:

http://127.0.0.1:5000

⚠️ Known Limitations

EXE size is large (expected due to Whisper + Torch)

First startup may take a few seconds

CPU-only Whisper (no GPU acceleration)

🔐 Security & Privacy

No internet connection required after build No data is uploaded anywhere All processing happens locally

Suitable for sensitive or confidential audio

🤝 Acknowledgements

• OpenAI Whisper
• PyTorch
• FFmpeg
• NLTK
• Flask
• PyInstaller

If you encounter issues:

Ensure you are using 127.0.0.1 and not a public IP Ensure the app is located in a writable folder Check console output if using a console-enabled build