Skip to content
EVENOR

Proof of concept

Early warning on students falling behind

The signs that a student is drifting are all present in systems the institution already runs: attendance dropping, internal marks slipping, assignments going in late or not at all. Nobody sees them together until the end of semester result makes it obvious, by which point the useful conversation is months late.

An AI agent node reading a task, choosing a tool and writing a result back

What we built

  • A view that reads attendance, internal assessment marks and assignment submissions together per student
  • Signals that surface a change in pattern rather than a low absolute number, so a steady student having a bad month is noticed
  • The reason stated in plain language for the mentor, naming the subjects and weeks involved
  • A weekly list for each mentor, ordered by how early an intervention would still help
  • No automated action on any student, ever. The system suggests a conversation and a person decides
  • Access limited to the mentor and the department, with the underlying data staying where it already lives

What it showed

The proof of concept surfaced students whose pattern had changed several weeks before the semester result would have shown it, and gave the mentor a specific reason to open the conversation with rather than a score.

What a live engagement needs

A live version needs read access to the attendance and assessment systems, the institution's own view of what counts as concerning, a clear position on who may see a flag, and agreement that no consequence for a student is ever automated.

Services behind it

Data and AI foundationsAI workflow automation
Talk about your own case

Related proof of concept work

See the whole library