AI-Powered Employee Attrition Prediction System

CH.Vasundhara 1, Jayanti Rama Rakesh 2

1Assistant Professor, 2 MCA Final Semester, Master of

Computer Applications, Sanketika Vidya Parishad Engineering College, Vishakhapatnam,

Andhra Pradesh, India

jayanthirakesh888@gmail.com

ABSTRACT:

Employee Attrition Prediction System is a Machine Learning-based application developed to predict whether an employee is likely to leave an organization. The system helps Human Resource (HR) departments identify employees with a high risk of attrition at an early stage. It analyzes various employee-related factors such as age, department, job role, monthly income, work-life balance, years at the company, overtime, and performance to generate accurate predictions. The application uses a React-based frontend, Node.js and Express backend, Flask API, and MongoDB database for efficient data management. A supervised Machine Learning model is trained on historical employee data to classify attrition risk. The system also provides risk scores and explainable insights using SHAP values, enabling HR professionals to understand the reasons behind each prediction. It supports both individual and bulk employee predictions for organizational use. The proposed system improves decision-making, enhances employee retention strategies, reduces recruitment costs, and increases workforce productivity. It offers a secure, user-friendly, and scalable solution for modern HR analytics. Overall, the system enables organizations to take proactive actions for retaining valuable employees and improving overall organizational performance.

Keywords : Employee Attrition Prediction, Machine Learning, Human Resource Analytics, React, Flask, MongoDB, SHAP, Classification Model, Workforce Planning, Employee Retention.