Quality & Accreditation · Continuous Evidence
Stop collecting evidence. Start capturing it.
The Evidence Library is the governed home for everything that proves you meet a standard. Evidence is captured where work happens and tagged to the criterion it proves as it lands, AI-assisted and human-confirmed. Every number stays bound to its document, so a review becomes a state you’re already in, not an event you scramble for.
Evidence Library
Tagging queue
Coverage by requirement
The shift
The scramble is a process problem, not an effort problem.
Weeks before a visit, the email goes out: “send me all assessment samples, marking schemes and minutes by Thursday.” What follows isn’t quality assurance. It’s a search party.
The marking scheme exists. The minutes were taken. The number is right. But the proof is three laptops away, because evidence is being collected after the fact, not captured as it’s made. Working harder won’t fix that, and a bigger shared drive won’t either.
The continuous-evidence model
Design evidence into the work, not into a project.
Four design choices turn evidence from something you gather into something your institution emits, built into normal work, not a pre-review fire drill.
Capture at the source
Evidence enters where the work is done — a marking scheme uploaded inside the course, minutes attached to the decision they record. The person closest to the work captures it correctly.
Tag to the criterion
As a document lands, the system proposes its tags — standard, KPI, course, term — and a person confirms. Assisted, not automated: a tagging policy that actually runs.
Bind the number to its proof
A KPI value and its evidence travel together. Documents reference the KPI engine; they never recompute a number. The figure always matches its proof.
Make coverage live
Coverage by criterion is a live view, not a spreadsheet rebuilt before each visit. You see which standards are backed and which are thin — long before a reviewer does.
When evidence is captured instead of collected, the review stops being an event you brace for and becomes a state you’re already in. The pre-visit email disappears. The dependency on the one person who knows where everything is disappears with it.
How evidence flows
A document's journey, from work to coverage.
Whether uploaded or authored, every artifact follows one path. What enters on the left becomes audit-ready evidence on the right, classified to the criterion it proves at every step.
Inside the product
One governed library, across its whole lifecycle.
The Evidence Library ingests, classifies, secures and surfaces evidence, both uploaded artifacts and authored governance documents, from capture to coverage. It is framework-neutral: configure the standards you report against — a regional mark like CAA, MQA or NAAC, plus international ones like ABET or AACSB — and the same capture-and-tag engine runs beneath whichever frameworks apply to you.
The searchable home of all approved evidence
Find evidence in plain English, or filter by type, tag, programme and term.
Confirm the suggested labels before filing
The system proposes tags with a confidence score; a person approves. Nothing auto-enters.
Quality control at the door
Faculty uploads pass a quick QA check. Approve to accept, or send back with a note.
Evidence by requirement, live
Which standards are backed and which are thin, computed live from the tags.
Authored with the real numbers built in
Authored with live data bound in, then published and frozen into the library.
Everyone sees only what they should
Public, private or restricted, with scoped guest access and a full audit trail.
Not another repository
How Continuous Evidence is different.
Most accreditation tools are a place you upload evidence into before a review. Continuous Evidence works the other way around — evidence accrues from daily work, tagged to the criterion it proves, in the framework you actually report against.
And there’s no process to bend to the tool: Continuous Evidence fits how your institution already works — the same tagged evidence serves every framework you report against.
The rules you can rely on
Assisted, not automated. Governed by design.
The Evidence Library does the heavy lifting (proposing tags, drafting narrative, surfacing gaps), but a person always owns the truth. These are the guarantees underneath every screen.
The system suggests; people decide
AI proposes labels and drafts. A person confirms. Nothing is filed or approved automatically.
Numbers can’t be faked
KPIs are computed by a deterministic engine. Documents only reference them, so the figure always matches its proof.
Published means permanent
Once published, a document is frozen forever: a dated record with its snapshot pinned. Only who can see it may change.
Access is enforced
Every document, search and download is scoped to the requester and enforced server-side, not by convention.
Everything is traceable
Uploads, tag changes, approvals, publishes and sharing are all recorded as an auditable trail of who did what, when.
Built for every role
One library. A different job for everyone who touches it.
A five-minute task in your own course. Upload the paper and marking scheme where you work. No framework to learn, no last-minute email.
See your programme’s coverage and chase only what’s genuinely missing. Confirm tags, review uploads, publish programme documents.
Prove any criterion on demand, and trust that every number matches its document. Coverage is live, not reconstructed under pressure.
Executives see real-time readiness to take to the board. External reviewers get a scoped, view-only slice: exactly their sample, nothing more.
What changes
What it looks like on a normal Tuesday.
Concretely, the week looks different, and you never sent an email.
A lecturer finishes grading and uploads the question paper and marking scheme inside their course, the way they’d naturally work. The system tags it to the assessment, the course, the term and the standard, and surfaces those tags for a quick confirm. It’s in the library, classified, in seconds.
A programme review that used to be written from scratch is generated with the live numbers already bound in: placements, graduates, outcomes. So the prose is judgement, not data entry. When it’s published it becomes a frozen, dated record that enters the library on its own.
When you want to know where you stand on a standard, you open a coverage view and see exactly which criteria are backed and which are not, by requirement, not by guesswork. The gap analysis you used to produce in a fire drill is simply there, current, all the time.
See it on your own frameworks
Watch continuous evidence run on your standards.
A 30-minute walkthrough on the framework you report against — CAA, MQA, NAAC, NBA, ABET or AACSB. Your criteria, your evidence, live coverage.
From the people who use it
Trusted for accreditation, in their words.
A named customer who used the Evidence Library to organise the documentation for accreditation, quoted with permission.
“The Creatrix Campus application has not only helped to evaluate our curricula; it has been key in organising the documentation necessary for accreditation. The team engaged with us regularly and no query or change was a problem; they understood our vision and enhanced the software to make it happen.”
Shared by a named customer, quoted in part with permission. More verified reviews on G2 and Capterra.
From the resources library
Go deeper: filtered to this product's tags.
The same tagging that organises your evidence organises our library. Filter by the tags that matter to Continuous Evidence and pull exactly what's relevant.
Frequently asked
Plain answers about the Evidence Library.
Does the AI decide what counts as evidence or how it’s tagged?
No. The AI proposes tags with a confidence indicator; a person reviews and approves. Nothing enters the library without someone confirming it.
The model is assist-tier: AI proposes, people decide. That’s what keeps classification consistent at thousands of documents without asking anyone to tag from a blank field.
Can the AI or the library change our KPI numbers?
Never. KPI values are computed by a deterministic engine. AI never authors a KPI, and documents only reference KPI values; they don’t recompute them.
Because the number and the proof are bound together by design, the figure in a report always matches the artifact behind it.
What happens to an authored document after we publish it?
It is frozen forever: an immutable, dated record with its data snapshot pinned and a PDF rendered, and it registers into the library through the tagging queue, where a person confirms which requirements it serves.
Sharing can change later; the content cannot.
Who can see a given piece of evidence?
Each item is public, private or restricted, with grants to departments, roles or specific people. QA and administrators always have access; the creator always retains theirs.
External and visiting reviewers receive a scoped, time-bound, view-only slice: exactly the evidence they need, never the whole institution.
Does this replace our shared drive or SharePoint?
It becomes the governed system of record for accreditation evidence: criteria-tagged, access-controlled and auditable. Personal drives stay for scratch work; what proves a standard lives in the library.
Will it work for frameworks beyond CAA, like MQA or NAAC?
Yes. The operational data model is framework-neutral; the standards taxonomy and KPI framework are the configurable layer. The same continuous-evidence engine applies across CAA, MQA, NAAC and others — only the set of standards changes.
Where does our data live and how is it secured?
In your own database, with row-level security and the access layer enforced server-side. The browser only ever holds a session token, never database or AI credentials.
And if the AI service is unavailable, uploading still works; documents are simply tagged by hand until it’s back.
Get started
If your last review ended with a search party.
The fix isn’t a bigger drive or an earlier email. It’s capturing evidence at the source and tagging it to criteria as it happens. Tell us where evidence breaks down today.