mock-mate

Introduction

MockMate is a comprehensive interview practice platform that simulates real-world technical interviews. The system combines AI-powered question generation, automated answer evaluation, and trending news-based interview questions to provide an up-to-date and realistic interview experience.

Setup

Prerequisites

• Python 3.10+ (3.12 is recommended)
• Node.js 18+ (20+ is recommended)
• npm 10+
• uv (if you don't have uv installed, you can install it with pip install uv)

Backend

cd backend
cp .env.example .env
uv venv
source .venv/bin/activate # might be different on your system, but uv will tell you the correct command
uv sync

Note: You should change the OPENAI_API_KEY in the .env file to your own.

For NEWS_FETCH_USER_AGENT in .env, you might not need to change it, but if the news fetching is not working, you can try to change it to your own browser's user agent. (Simply google "my browser user agent")

Run in development mode

./run.sh # The port is 5200 by default, you can change it in the script

API Documentation

Visit http://localhost:5200/docs to view the API documentation.

Frontend

cd frontend # if you are in the root directory
cp .env.example .env
npm i

Note: You can change the values in the .env file to your own.

Run in development mode

The port is 1314 by default, you can change it in vite.config.ts

npm run dev

Production Update Script

For the Ubuntu server deployment flow, this repo provides a one-click update script: update.sh.

It is designed for the deployment layout described in DEPLOYMENT.md (/root/mockmate/app) and will:

• Pull latest code (git pull)
• Sync backend dependencies (uv sync)
• Restart backend service (systemctl restart mockmate)
• Rebuild frontend assets (npm install and npm run build)

Usage on server:

cd /root/mockmate/app
chmod +x update.sh
./update.sh

Notes:

• Run as a user with permission to restart systemd services (usually root)
• Requires the same conda/nvm environment paths used in DEPLOYMENT.md

Database Maintenance

The system automatically fetches news and generates interview questions on a schedule. Over time this data accumulates. A scheduled task runs every 24 hours to clean up expired data (default retention: 30 days). You can also run the cleanup manually:

cd backend

# Preview what would be deleted (no changes made)
python cleanup_news.py --dry-run

# Clean up with default 30-day retention
python cleanup_news.py

# Custom retention period
python cleanup_news.py --days 7

To change the default retention period, set the NEWS_RETENTION_DAYS environment variable in your .env file:

NEWS_RETENTION_DAYS=14

The cleanup process:

1. Deletes expired NewsBasedQuestion and associated Question records
2. Skips questions that have been used in user interview sessions
3. Deletes expired NewsItem records
4. Runs SQLite VACUUM to reclaim disk space

System Architecture (Task Design Document)

See MockMate - Interview Simulation System Architecture

For more fancy documentation, please visit this repo's DeepWiki !