All projects
08 / Apr 2026/ MS final project, Governors State University, team of 3
AI-Driven Supply Chain Risk Predictor
Planners find out a shipment is late only after it is late.
- Role
- One of three engineers on the team.
- Stack
- Python, scikit-learn, Streamlit
- Context
- MS final project, Governors State University, team of 3
01
What I built
- 01A web app that predicts shipment delay in days so planners can flag risky shipments before they ship.
- 02Compared Linear Regression, Random Forest and Gradient Boosting in scikit-learn; deployed Gradient Boosting.
- 03Three tabs: single-shipment prediction, batch CSV upload and model analytics.
02
Software
- 01Python, pandas
- 02scikit-learn
- 03Streamlit, deployed live
03
Validation
- 01Three regressors compared on MAE and RMSE; Gradient Boosting performed best and was deployed.
04
Results
- 01Live demo presented at the final program review, April 2026.
05
Characteristics
| Parameter | Value |
|---|---|
| Deployed model | Gradient Boosting |
| MAE | 0.422 days |
| RMSE | 0.533 days |