Workforce Insight
End-to-end attrition prediction system with an explainable Flask API for HR retention.
- CLIENT
End-to-end attrition prediction system with an explainable Flask API for HR retention.
- SCOPE
End-to-end attrition prediction system with an explainable Flask API for HR retention.
- RELEASE
End-to-end attrition prediction system with an explainable Flask API for HR retention.
- WEBSITE
- https://github.com/dimplebhardwaj2/Workforce-Insight
- CREDIT
- Producer
- AI / ML Engineer
- Art Director / Designer
- Machine Learning
- Illustrator
- Python, PyTorch, Scikit-learn, Flask, Pandas
- Frontend Engineer
- 2025
End-to-end attrition prediction system with an explainable Flask API for HR retention.
Predict, explain, and retain
Built an end-to-end Employee Attrition Prediction System, benchmarking Logistic Regression, Random Forest, Gradient Boosting, and a PyTorch MLP; achieved 68% recall for at-risk employees vs. 8% for the tuned baseline.
Engineered a 49-feature ML pipeline with class-weighted loss for imbalanced data (~5:1), 5-fold cross-validation, and GridSearchCV, achieving 0.82 ROC-AUC.
Developed a Flask REST API serving real-time predictions with counterfactual explainability, converting model outputs into actionable HR retention recommendations.
End-to-end attrition prediction system with an explainable Flask API for HR retention.
https://github.com/dimplebhardwaj2/Workforce-Insight
