Student retention
Dropout, cohorts, early warning with a predictive model, and tracking of the interventions that help more students finish their degree.
The challenge
Universidad del Ceibo learned about dropout after it happened. It needed to see each cohort’s trajectory, spot in time who to support, and know whether its retention programs actually reduce dropout.
Python (modelo predictivo) · Power Query · DAX · Power BI
Key findings
- Dropout per term fell from 23.7% (2021-1) to 7.1% (2026-1); the first year concentrates 71% of dropouts.
- 727 students are at high risk, 245 still without intervention; the most common factor is a low GPA (< 3.0).
- Among high-risk students, 19.0% of those supported drop out vs. 27.2% without support (8.2 pp less).
Capabilities shown
- Predictive model (logistic regression, AUC 0.75)
- Actionable early warning with suggested intervention
- Cohort survival and first-year retention
- Intervention evaluation
Dashboard tour
6 views. Click any view to enlarge it.
Live interactive dashboard
Use the filters and the navigation menu inside the dashboard, just as a manager would.
More prototypes from this service
n8n + AI automationAI admissions assistant
Every applicant gets a personalized reply within minutes, and advisors know who to call first.
View prototype
n8n + AI automationDropout early warning
Every Monday, at-risk students get support before it’s too late.
View prototype
n8n + AI automationHelp desk agent with RAG
An agent answers 24/7 from the official rules and escalates to a person when needed.
View prototype
AI agentAI academic tutor
A Socratic tutor that explains with the course material, generates exercises, gives hints without revealing answers, and feeds a teacher dashboard.
View prototypeReady to turn your data into smart decisions?
We build this kind of solution with your own data. Let’s start with a free 30-minute call.