The diagnostic that maps a learner in twenty-four questions.
Placement exams are long because fixed tests waste most of their questions on the wrong difficulty. An adaptive diagnostic chooses each item for maximum information about this learner — which is why it converges on an honest starting map in roughly the time a coffee takes.
What the diagnostic is actually doing, item by item
Under the hood is item response theory — the psychometrics behind serious standardized testing. Every question in the bank carries calibrated difficulty and discrimination parameters; the learner carries a live ability estimate with an uncertainty band. Each next item is chosen where it will shrink that uncertainty fastest — right at the estimate’s current edge, where the answer is genuinely informative.
Answer by answer, the band collapses. Twenty-four well-chosen items typically pin the estimate as tightly as a hundred fixed ones, because none are wasted confirming the obvious. And since practice keeps updating the same model afterwards, the diagnostic isn’t an event — it’s just the model’s fast start.
A starting map, not a placement score
The output feeds straight into practice: per-concept levels across the knowledge map, so day-one sessions already skip the known and target the gaps.
Calibrated items, continuously audited
New questions enter with prior parameters and calibrate against live data; drifting or ambiguous items flag themselves for review. The instrument stays sharp because it measures itself.
Honest about its own uncertainty
Estimates ship with their confidence bands, and thin evidence looks thin on every dashboard. A diagnostic that overstates its precision is just a slower way to guess.
Twenty-four questions to a map
The diagnostic narrows from wide uncertainty to a level in one sitting — each question chosen by the last answer.
Interface shown as an illustration with representative numbers, not a screenshot — the layout is the product’s.
Sit the diagnostic yourself.
The fastest way to trust it: take one in the demo and watch your own estimate converge, item by item.
The evidence this page stands on
Questions buyers ask
Is 24 questions really enough?
For a working starting estimate per path, yes — that’s what maximum-information selection buys. The estimate then keeps sharpening through ordinary practice, so precision compounds instead of expiring.
How do learners experience it?
As a short, oddly well-pitched quiz — items sit near their edge by design, so it feels challenging but fair. No time pressure theatrics; the information comes from the answers.
What stops the diagnostic being gamed?
Deliberately tanking it only earns easier practice that the engine then adjusts upward within days as real performance shows. There’s no lasting advantage to fooling your own starting point.
Do you re-run diagnostics periodically?
Rarely needed — practice data keeps the model current. Fresh diagnostics fire for genuinely new territory: a new path, a role change, a long absence.
Can we use it standalone, as an assessment product?
Its job here is to start and steer learning. For standalone selection-grade assessment you want an assessment platform — a different product discipline with its own proctoring and fairness machinery.
See it on your own content.
Bring one course. We’ll show you the retention curve your current training leaves behind — and what scheduled review does to it.
- 30 minutes, on your calendar — pick a slot here
- Run on your own content wherever possible, not a canned deck
- You see the dashboards, the learner surface and the evidence exports
- No commitment — and pilot data stays yours either way