Skip to content
Project number
Project / 008
Year
2025
Category
Analytics
Status
Completed · Live

HR Analytics & Attrition Model

Predict employee attrition with ML models and an interactive dashboard for retention insights.

Stack
Python · pandas · scikit-learn · Streamlit
Tags
Predictive Modelling · HR Analytics

01 /Overview

Organisations want to know which employees are at risk of leaving, and why, before it happens.

02 /What I built

  • Exploratory analysis of department-level attrition trends
  • Feature engineering, scaling and encoding
  • Multiple models compared, including Logistic Regression and Random Forest
  • Interactive dashboard for exploring the data and getting predictions
  • Findings: overtime raises quit rates; work-life balance often matters more than salary; growth opportunities reduce turnover

03 /Outcome

A live Streamlit dashboard on the IBM HR Analytics Attrition dataset that explores attrition drivers and predicts attrition risk.

04 /What I learned

The most useful output was not the prediction but the drivers: overtime, work-life balance and career growth explained more than salary did.