Decide intervention per at-risk learner
Packaged know-how that tells an agent how to do a job well.
You might say…
“The tool gave me a list of 11 flagged learners. I don't have bandwidth to call all of them. I need to know which three actually matter.”
What it does
Given the computed risk list plus context on each flagged learner, judges who genuinely needs an intervention now versus who can wait, and what kind.
Trigger: Use after the risk list is computed to convert flags into a specific, actionable intervention plan — not every flag warrants the same response.
I/O: Ranked at-risk list + per-learner context → intervention plan (who, what kind, when) with reasoning for each decision
Recognise the problem?
The primitives are the commodity part. The fastest next step is a conversation about composing them into something that works for you.
Start a conversation