AI Copilot

AI Interview Copilot is a desktop assistant built with Python that helps users handle live interview scenarios. It records microphone input, transcribes it in near real-time using Whisper (via faster-whisper), and streams the recognized text to OpenAI GPT models to draft structured interview responses in different styles.

✨ Features

• 🎤 Real-time microphone recording and transcription (Whisper).
• 🤖 AI-powered answer drafting with OpenAI GPT (streaming responses).
• 🖥️ Tkinter-based desktop UI with live transcription preview.
• 🔒 Window protection against screen sharing on Windows (via SetWindowDisplayAffinity).
• 📋 Copy-to-clipboard and clear text utilities.
• ⚙️ Configurable Whisper model sizes and GPU/CPU selection.
• 📝 Detailed logging in both console and logs/interview_copilot.log.

📦 Requirements

• Python 3.10+
• sounddevice
• numpy
• openai
• faster-whisper
• python-dotenv
• tkinter (pre-installed with most Python distributions)

🚀 Usage

1. Clone the repository: git clone https://github.com/SteveParadox/Ai-Copilot.git

2. Navigate into the project folder: cd Ai-Copilot

3. Install dependencies: pip install -r requirements.txt

4. Create a .env file in the project root with your OpenAI API key:

5. OPENAI_API_KEY=your_api_key_here
   

6. Run the application: python ai_copilot.py

7. Click Record to start transcribing speech and generating interview answers.

---

📝 Logs are saved in the logs/ directory. For troubleshooting, check interview_copilot.log.