"Prove you can do this."
The seven topic pages teach techniques in isolation. These six cases ask you to integrate them. Each is a fictional but realistic company facing a real decision — fraud, maintenance, demand planning, recommendations, segmentation, optimization. Each case kit includes the dataset, the analytical notebook, an Excel scoring tool you'd hand to operations, a student worksheet, and one or more sample executive presentations.
How to use these
Pick the case closest to your group project's domain. Work through the notebook, fill in the worksheet, then re-create the executive deck in your own words. Whether you use Python, SAS, or Excel for the modeling is up to you — the deliverable is the decision, not the tool.
The six cases
Optional reference: how models reach production
What happens after you've trained a model and want to put it to work — and the human skills you'll need around it.
- How Models Reach ProductionFrom notebook to scoring service — the practical path to deployment.
- Operationalizing Unsupervised ModelsSpecial challenges of putting clustering and anomaly detection into production.
- SAS Viya Deployment GuideIf you're going the enterprise route, the full SAS deployment playbook.
- Model Deployment Guide (interactive)Browser-friendly walkthrough of the deployment lifecycle.
- TrendCart Deployment ChecklistA concrete worked checklist for a retail deployment.
- Deployment hubLanding page collecting all the deployment material.
- Deployment types — batch, real-time, edgePick the right architecture for the use case.
- Analyst Soft-Skills Self-AssessmentA self-check on the human skills that complement the technical ones.
- Skills in the AI AgeWhich analyst skills hold their value as AI gets stronger.