Clear process. Rule-based checks. Human judgement.

How Ref-Check works

A simple, explainable workflow from upload to report.

Ref-Check is designed to be clear, predictable and easy to use. It checks citations and reference lists through structured, rule-based processes so users can review findings without needing technical expertise.

No generative AI or machine-learning models.

Ref-Check uses deterministic logic, trusted metadata sources and explainable checks. Human academic and professional judgement stays central.

The workflow

From upload to report in three clear steps.

Ref-Check provides structured information about citation and reference accuracy. It does not judge the quality of the work or make decisions for the user.

1

Upload a document

Upload written work that includes a reference list. Ref-Check reads the document structure and separates the main text from the references where possible.

2

Run the checks

Ref-Check analyses citations, reference entries, key identifiers such as DOIs and URLs, and the selected referencing approach.

3

Review the report

Users can review grouped findings, check individual references, and use the report to support feedback, tutorials, quality review or follow-up discussion.

What happens behind the scenes

Ref-Check looks for patterns that take time to spot manually.

The checks are designed to make the reference trail easier to see, while leaving interpretation and action to the person reviewing the work.

Citations

In-text citation detection

Finds author–date citations and compares them with the reference list.

References

Matching and gaps

Highlights citations with no matching reference and references that do not appear to be cited.

Style

Formatting indicators

Flags common structure and formatting issues against the selected referencing approach.

Verification

Source checks

Uses trusted bibliographic and metadata sources to indicate where details may need closer review.

Designed to stay explainable

Transparent checks, clear limits and secure handling.

Ref-Check is built to support review rather than replace it. Its role is to surface useful information so users can make informed academic, professional or quality decisions.

No AI models

Not generative AI

Uploaded work is not processed by external generative AI or machine-learning services.

Rule-based

Predictable logic

Checks follow deterministic, explainable processes rather than opaque judgement.

Secure

Stored for review

Content is stored securely to support reporting, user access and review.

Human-led

Judgement stays central

Ref-Check highlights findings; users decide what those findings mean in context.

Ready to see how Ref-Check works?

Start a check, explore the features, or use the Learning Hub for plain-English referencing guidance.