Estimates from the institution's own data
A machine-learning model retrained regularly on the institution's own quiz and assignment scores, attendance and engagement data. Until there is enough data to train it, the system uses published rules.
The system estimates each student's need for support from their grades, attendance and engagement with course content, and shows the factors that raised it and the weight of each. When the level goes up, the teacher and the parent are notified.
A machine-learning model retrained regularly on the institution's own quiz and assignment scores, attendance and engagement data. Until there is enough data to train it, the system uses published rules.
Each alert lists the factors that raised the level and the weight of each, calculated with SHAP and LIME and written so teachers and parents can follow them.
Teachers see their students ordered by level of need, administrators see how levels are distributed and which factors come up most, and parents see their child's status in their portal.
When a student moves to the high level, the teacher is notified in the platform and by email, and the parent through the enabled channels.
The early-warning dashboard runs inside Blackboard, Moodle and Canvas through LTI 1.3 without moving any courses, and the student list exports to Excel.
See which students need support and why, and decide where to start.
See the level of need and the reasons, next to attendance and grades.
Follow how levels are spread across the institution and which factors recur most.
The products run on one database, so quiz and attendance results appear in the gradebook, the parent portal and early warning without being entered again.
Academic administration
Attendance, gradebook, reports and permissions.
DetailsCommunication and parent portal
A parent portal, with notifications in the platform, by email, by SMS and on WhatsApp.
DetailsAssessment
A question bank with eight question types, and suggested grades the teacher approves.
DetailsNo. It raises alerts and explains them; the educational decision stays with the teacher.
Previous assessment results and their trend, the previous failure rate, gaps in assessment activity, attendance over the last 30 days, and engagement with course content.
No. The model service runs as part of the platform itself.
The ClassWhy team will show you the platform for your type of institution and answer your setup and migration questions. Please request a demo and the team will contact you to arrange a time.