Methodology
Where the rubric comes from, and where it stops.
Rubricare's review process in four steps: what it reads, what it checks against, what it hands back, and what stays a clinical decision throughout.
The plan is read in full
A completed or draft Behaviour Support Plan is read section by section, preserving the structure of the document rather than treating it as a block of undifferentiated text.
Every section is checked against the rubric
The rubric is informed by the BSP-QEII instrument, used here as a positioning anchor rather than a live, publicly reported score. Each item on the rubric maps to a specific, named criterion, not a general impression of quality.
Flags are traced back to the plan text
Where the review raises an issue, it is linked to the exact passage that triggered it. A practitioner can check the reasoning against the source in seconds, rather than having to take a summary judgement on trust.
A clinician decides what happens next
The review is handed back as a set of flagged passages, not a verdict. Nothing about accepting, rejecting, or acting on a flag is automated. That decision, and the clinical responsibility that comes with it, stays with the practitioner.
On BSP-QEII
BSP-QEII is used here as the framework that gives Rubricare's rubric its structure and its vocabulary. It is a positioning anchor for the review, not a live, publicly reported score, and Rubricare carries no NDIS Commission affiliation or endorsement.
No accuracy, agreement, or time-saved figures are published while the validation study is underway. Anything stated about how well Rubricare performs will be published once that study closes, not before.
Data handling, in brief
Plan content is processed to generate a review and is not used to train or improve any underlying model. Production data resides in Australia. Where plan content is sent to a processing service outside Australia, that cross-border step is disclosed in the privacy policy.