Conversational AI for Technical Interview Practice

A web-based platform that uses conversational AI to help Computer Science students practice the think-aloud process during technical interviews. This system was developed as part of the research presented in the paper "Designing Conversational AI to Support Think-Aloud Practice in Technical Interview Preparation for CS Students" published at IEEE VL/HCC 2025.

• Paper: arXiv:2507.14418
• Conference: IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC) 2025

🎯 Overview

Technical interviews are crucial for CS students, but the think-aloud process—where candidates verbalize their thought processes while solving coding problems—remains challenging to practice. This platform addresses this gap by providing an LLM-based conversational AI system that offers:

• Techincal Interview Simulation: AI-facilitated mock interview tool that simulates a technical interview experience through voice-based natural conversation
• AI Feedback on Think-Aloud Practice: Providing feedback on users’ think-aloud performance based on the interview simulation (A) transcript
• AI-Generated Think-Aloud Example Dialogue: Providing (A) AI-generated dialogues that model thinking aloud during a technical interview for each coding problem. The goal is to help users learn from examples relevant to their current task and better articulate their thought processes

🛠️ Technology Stack

Backend

• Python Flask - Web framework with async support
• OpenAI API - Large Language Model integration
• MongoDB (Optional) - Database for storing sessions and user data
  • Can run without MongoDB using in-memory storage for development/testing

Frontend

• Vue.js 3 - Progressive JavaScript framework
• TypeScript - Type-safe JavaScript development
• CodeMirror 6 - Advanced code editor with syntax highlighting
• Bootstrap 5 - Responsive UI framework
• Axios - HTTP client for API requests

📋 Prerequisites

• Python 3.8+
• Node.js 18+
• OpenAI API Key (Required)
• MongoDB (Optional - for persistent storage. Without it, data is stored in-memory and lost on restart)

🔧 Installation & Setup

1. Clone the Repository

git clone https://github.com/your-username/coding_interview_practice_platform.git
cd coding_interview_practice_platform

2. Backend Setup

cd backend

# Install Python dependencies
pip install -r requirements.txt

# Set up environment variables
cd src
cp .env.example .env
# Edit .env file with your API keys and database URLs
cd ..

Configure the backend .env file (backend/src/.env):

Option 1: With MongoDB (Recommended for Production)

MONGO_URI=mongodb+srv://username:password@cluster.mongodb.net/database_name
OPENAI_API_KEY=your_openai_api_key_here
OPENAI_GPT_MODEL=gpt-4o-mini
JWT_SECRET_KEY=your_jwt_secret_key_here (can be any random string - just type random words!)

📚 Need help setting up MongoDB? Check out the MongoDB Atlas Tutorial for easy cloud database setup (free tier available).

Option 2: Without MongoDB (Development/Testing)

# Leave MONGO_URI empty or comment it out to use in-memory storage
MONGO_URI=
OPENAI_API_KEY=your_openai_api_key_here
OPENAI_GPT_MODEL=gpt-4o-mini
JWT_SECRET_KEY=your_jwt_secret_key_here (can be any random string - just type random words!)

Note: When running without MongoDB, interview transcripts and feedback are stored in memory and will be lost when the server restarts. This mode is suitable for development and testing only.

3. Frontend Setup

cd frontend

# Install Node.js dependencies
npm install

# Set up environment variables
cp .env.example .env
# Edit .env file with your backend API URL

Configure the frontend .env file (frontend/.env):

VITE_BACKEND_URL=http://127.0.0.1:5000

🚀 Running the Application

Development Mode

1. Start the Backend Server:

   cd backend
   flask --app src/app run
   # Server runs on http://127.0.0.1:5000

2. Start the Frontend Development Server:

   cd frontend
   npm run dev
   # Application runs on http://localhost:5173

Production Build

1. Build the Frontend:

   cd frontend
   npm run build

2. Deploy Backend:

   cd backend
   # Use the provided deploy.sh script or deploy manually
   gunicorn --bind 0.0.0.0:5000 src.app:app

📁 Project Structure

coding_interview_practice_platform/
├── backend/                    # Flask backend application
│   ├── src/
│   │   ├── app.py             # Main Flask application
│   │   └── routes/            # API route implementations
│   ├── requirements.txt       # Python dependencies
│   ├── Dockerfile            # Docker configuration
│   └── deploy.sh             # Deployment script
├── frontend/                  # Vue.js frontend application
│   ├── src/
│   │   ├── App.vue           # Main Vue application
│   │   └── components/       # Vue components
│   ├── package.json          # Node.js dependencies
│   └── vite.config.ts        # Vite build configuration
└── public/                   # Static assets and documentation images
    ├── 1_interview simulation.png
    ├── 2_feedback.png
    └── 3_example.png

📝 Citation

If you use this platform in your research or find it helpful, please cite our paper:

@inproceedings{daryanto2025designing,
  title={Designing Conversational AI to Support Think-Aloud Practice in Technical Interview Preparation for CS Students},
  author={Daryanto, Taufiq and Stil, Sophia and Ding, Xiaohan and Manesh, Daniel and Lee, Sang Won and Lee, Tim and Lunn, Stephanie and Rodriguez, Sarah and Brown, Chris and Rho, Eugenia},
  booktitle={2025 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)},
  year={2025},
  organization={IEEE}
}

📞 Contact

For questions about this research or the platform, please contact:

• Taufiq Daryanto - Email

🔗 Links

• Paper: arXiv:2507.14418
• Conference: IEEE VL/HCC 2025

---