International Journal of Science, Engineering and TechnologyAuthors: Ms. Akanksha Patil, Aryan Gole, Harsh Pokharkar, Sanika Surve
Abstract: Due to the rise of remote work, ensuring employee accountability and performance has become increasingly challenging. This has also increased cases where employees misuse work hour using keyboard jigglers, auto-clickers, or embedded chips that simulate activity, allowing employees to appear active without actually working. This project presents an AI-driven system to detect such fraudulent behavior by analyzing keystroke dynamics, mouse movement patterns, and system activity logs. The goal is to build this system to detect such behavioural data including typing speed, key press duration, inter- key delay, mouse trajectory, and interaction timing to build a unique profile for each user. These behavioural patterns are then analyzed using machine learning algorithms to detect anomalies that indicate potential fraudulent activity. The System can keep the track of what people are actually doing on their devices in real time like how someone types or moves the mouse, the system can tell the work is being done by real person or just by using an automation tool. This project aims to improve transparency in remote job roles and promote honest work culture.
