VoiceFlow

VoiceFlow is a background audio transcription and processing tool for Ubuntu that allows users to quickly transcribe and process their speech using a hotkey combination.

Installation

1. Clone this repository: git clone https://github.com/yourusername/voiceflow.git cd voiceflow Copy

2. Create a virtual environment and install dependencies: python3 -m venv venv source venv/bin/activate pip install -r requirements.txt Copy

3. Set up your API keys: Edit voiceflow.py and replace the placeholder API keys with your actual keys for Groq, Deepgram, and GPT-4.

4. Set up the systemd service: sudo cp voiceflow.service /etc/systemd/system/ sudo systemctl daemon-reload sudo systemctl enable voiceflow.service sudo systemctl start voiceflow.service Copy

Usage

1. Press and hold Super + 3 to start recording.
2. Release Super + 3 to stop recording and process the audio.
3. The processed text will be copied to your clipboard and inserted into the active window.

Features

• Hotkey activation (Super + 3) for audio recording
• Maximum recording duration of 5 minutes
• Automatic transcription using Groq API (fallback to Deepgram Nova 2)
• AI processing of transcription using Groq's Llama model (fallback to GPT-4)
• Visual indicator for recording status
• Autostart on system boot

Areas for further development or refinement

• Improving the visual indicator (e.g., using custom icons or animations).
• Implementing a method to insert text into the active window (this can be challenging and may require additional libraries or system-specific methods).
• Adding more robust error handling and logging.
• Implementing a configuration file for easy customization of API keys and other settings.

Configuration

VoiceFlow uses environment variables for API keys. Set them in your shell or add them to your .bashrc or .bash_profile:

export GROQ_API_KEY=your_groq_api_key export DEEPGRAM_API_KEY=your_deepgram_api_key export OPENAI_API_KEY=your_openai_api_key

Permissions

VoiceFlow requires access to the microphone and the ability to write temporary files. The script will check for these permissions on startup. If you encounter permission issues, ensure that your user has the necessary rights to access the microphone and write to the temporary directory.

Tray Icon Support

VoiceFlow attempts to create a tray icon for easy access. If your system doesn't support tray icons, the application will still run but without a visible icon. You can still use the hotkey functionality.

Setting up VoiceFlow as a System Service

After making changes to the VoiceFlow script, follow these steps to set it up as a system service:

1. Create a systemd service file:

   sudo nano /etc/systemd/system/voiceflow.service

2. Add the following content to the file (replace yourusername with your actual username and add your API keys):

   [Unit]
   Description=VoiceFlow Audio Transcription Tool
   After=network.target
   
   [Service]
   ExecStart=/home/yourusername/Documents/Projects/Voiceflow/venv/bin/python /home/yourusername/Documents/Projects/Voiceflow/voiceflow.py
   Environment=DISPLAY=:1
   Environment=XAUTHORITY=/home/yourusername/.Xauthority
   Environment=GROQ_API_KEY=your_groq_api_key
   Environment=DEEPGRAM_API_KEY=your_deepgram_api_key
   Environment=OPENAI_API_KEY=your_openai_api_key
   Restart=always
   User=yourusername
   
   [Install]
   WantedBy=multi-user.target

3. Save and exit the editor.

4. Reload the systemd daemon:

   sudo systemctl daemon-reload

5. Enable the service to start on boot:

   sudo systemctl enable voiceflow.service

6. Start the service:

   sudo systemctl start voiceflow.service

7. Check the status of the service:

   sudo systemctl status voiceflow.service

8. To view logs:

   sudo journalctl -u voiceflow.service

After completing these steps, VoiceFlow will run as a system service and start automatically on boot. You can use the hotkey combination to activate it without needing to manually start the script.