A teacher with twelve students knows roughly who is struggling. A teacher with a hundred and twenty does not, and no amount of dedication fixes that - it is an arithmetic problem, not a caring problem.
The usual answer is a class average, which is the one number guaranteed to describe nobody in the room.
One picture of each student, fed from everything
ECMS keeps a mastery figure for each student on each topic, and it is fed from three places at once: marked exams, auto-graded assignments, and spaced review sessions.
That last point is the one that matters commercially. Most platforms have a tutor that knows nothing, a quiz that knows nothing, and a dashboard that knows a little - three features that never talk. Here they read and write the same record, so the picture gets sharper every week the centre operates, from work the students were doing anyway.
What the centre can do with it
The tutor teaches from where the student is. Not from where the syllabus assumes they are. Same AI tutor, different starting point per child.
Practice aims at the gap. Questions come from the centre's real question bank, targeted at the student's weakest topics, with difficulty adjusting from their last few answers. Every answer feeds back into the picture, so practising makes the model better, not just the student.
Review arrives before forgetting does. Spaced repetition schedules each fact per student, so a child revisits what they are about to lose rather than what the weekly plan happened to list.
The warning comes early. A falling mastery trend on the topics the next exam covers is visible before the exam, not after it.
A competitor can ship a feature list in a quarter. What they cannot ship is two years of records of what every student in your centre got right and wrong. The longer a centre runs on it, the more it is worth - which is the opposite of how most software ages.
The honest limits
The picture is only as good as what feeds it. A centre that marks on paper and never enters the results has no learner model, and no amount of AI invents one.
It also does not replace teaching. It tells a teacher which four students to spend Thursday on, and what to spend it on. Deciding how to explain a thing to a fifteen-year-old who has decided they are bad at maths is still the hardest part of the job, and still a person's job.
What the system removes is the guessing about who needs the help.