Create
Author by hand. Import what you have. Use AI as a drafting assistant — your choice.
Author items in NUADU's editor, import from your existing question bank, or use AI as an opt-in drafting assistant for candidate items. Whichever path you choose, every item that reaches a learner has been reviewed and approved by a named educator and is aligned to your rubric and framework.
Switch on AI assistance and it drafts candidate items for you to review, edit and approve.
Nothing enters the bank unreviewed.
Authoring takes judgement. NUADU shortens the drafting, not the judgement.
A good assessment item reflects what an educator wants to test, how they want to test it, and what they consider sufficient evidence of understanding. That judgement is the assessment's value. It cannot be outsourced.
What can be shortened is the drafting work — the typing, the formatting, the first-pass framework alignment, the first draft of distractors, the first draft of a rubric. Some teams want AI drafting on; some teams want it off. NUADU treats that as a configuration choice.
Author your way. AI is there if you want it. The educator decides what reaches the learner.
Three paths to a ready assessment.
Author or import
Author items in NUADU's editor, or import from your existing question bank — including items already aligned to your standards. AI is not required at this step.
Optional — AI drafting assistant
If your institution enables it, AI can draft candidate items from your materials — readings, training documents, course notes. It proposes question type, level and a draft rubric. Every candidate item is held for educator review.
Educator review and approval
The educator reviews each item — manually authored, imported or AI-drafted. Edit, reorder, replace, reject. Adjust rubric weights. Confirm framework alignment. Approve for publication. Every published item carries a named approver and an audit record.
Capabilities that respect authoring judgement.
AI drafting (opt-in)
When enabled, AI drafts candidate items from your materials. Educator reviews and approves every one.
24+ question types
MCQ, constructed response, spoken response, drag-and-drop, hotspot, matching, ordering, labeling, code.
Framework alignment, as a proposal
Suggested alignment to Bloom's, CEFR, EQF or your custom framework. Educator confirms.
Rubric-anchored authoring
Every item is connected to your rubric. AI-drafted rubrics are starting points; the educator finalises.
Rich media
Images, audio, video, simulations and interactive elements — for items that need them.
Item bank with audit trail
Versioned items, named approvers, change history. Ready for QA and external review.
Authoring for how educators actually work.
A weekly course quiz
Educator imports their slide deck. AI proposes 15 candidate items. Educator reviews, edits three, rejects two, publishes the rest.
- From slides to candidate items
- Educator reviews every item
- Named approver on each
Corporate compliance refresh
A regulation update changes 12 paragraphs. L&D feeds the update to NUADU. AI proposes new candidate items and retires items now out of date. Compliance lead reviews and signs off. The audit trail records what changed and who approved.
- Regulation-aware drafting
- Retirement proposals
- Compliance sign-off and audit record
Certification item development
A certification body builds its item pool using item banking with calibration statistics. Subject-matter experts draft and review items; AI can be used as a first-pass drafting aid. Items go through psychometric review, calibration statistics and equating before they enter the live pool.
- SME-led authoring
- Optional AI first-pass
- Psychometric review and calibration statistics
Language proficiency battery
A language testing programme assembles items across reading, writing, listening and speaking. Speaking and pronunciation items are authored alongside the rubrics that will be used to assess them. CEFR-referenced.
- Authored alongside speaking rubrics
- CEFR-referenced
- Inter-rater reliability tracked from day one
Why this is authoring, not item generation.
Author your way
Manual, imported, or AI-drafted. Institutions choose which. AI is opt-in per institution, per programme, per assessment.
AI never publishes
Every published item has been reviewed and approved by a named educator. AI is a drafting assistant. The educator is the author of record.
Framework alignment is a proposal
AI proposes a framework alignment. The educator confirms or changes it. The proposal does not become the alignment until a human says so.
Rubric-anchored from the start
Items are connected to the rubric criteria they assess. That rubric is what the marking step uses later — same rubric, end to end.
Audit-ready by default
Every item has a version history, a named approver, and an audit record. Ready for QA review, external examination and accreditation.
Multilingual
Authoring, alignment and marking in multiple languages. Translations reviewed by named educators — not auto-published.
Questions, answered.
Try authoring your way.
Author manually, import what you have, or switch on AI drafting. Either way, the educator decides what reaches the learner — and every item carries a named approver.