CollabCode: A Real-Time Collaborative Coding Platform with Artificial Intelligence Code Assistance and Secure Multi-Language Execution

This work presents CollabCode, a browser-based platform for collaborative programming. It combines shared code editing,artificial intelligence assisted coding, server-side execution, persistent sessions, and a Secure Interview Mode in one environment. The system is implemented using React.js, Node.js, Socket.IO, MongoDB, Monaco Editor, and Docker. Editor changes are transmitted through persistent WebSocket connections, while AI requests use a separate communication path so coding activity can continue without interruption. Programs written in C, C++, Java, and Python are executed in isolated containers with resource and network restrictions. Users can create coding rooms and recover stored sessions. In the conducted tests, synchronization remained below 50 ms for rooms with up to ten concurrent users, execution responses remained below 1.5 seconds, and message delivery reliability stayed above 98%. Secure Interview Mode records browser and session events and provides the interviewer with a security view. The platform is intended for collaborative programming, pair programming, and technical interviews.