AEWIS: Academic Early Warning & Intervention System

In traditional academic environments, "at-risk" students are often identified only after they have failed their final exams. By this point, it is too late for intervention. AEWIS was designed to close this gap by analyzing attendance and mid-term data in real-time, providing educators with a "predictive window" where they can actually make a difference.

Most educational institutions are hesitant to adopt new software because of the friction involved in integrating with existing ERP systems. I designed AEWIS with a Zero-ERP Integration strategy. Instead of complex API hooks, the system allows admins to simply upload existing CSV/Excel exports from systems like Fedena or Moodle.

The "Risk Engine" is built using Vectorized Pandas logic. By avoiding iterative loops and using vectorization, the engine can scan through thousands of student records instantly to flag students based on three primary metrics: Attendance (below 75%), Trend Analysis (score decline of 15%+ across assessments), and Absolute Risk (below 40% in core subjects).

Using Streamlit and Plotly, I created "Admin Heatmaps" that visualize "crisis subjects"—classes where a large percentage of students are flagging as at-risk. Teachers have an "Intervention Loop" where they can log actions and see the student's risk score update in real-time.