Similarity and AI-writing signals guide the review — the decision remains human.See how it works

How it works

A complete review workflow, not a black-box score.

Every stage is designed to help authorised people understand the evidence and make a proportionate decision.

1. Secure submission

Upload PDF, DOCX or TXT files up to 50 MB. Large files go directly to private storage and are processed asynchronously.

2. Data minimisation

Configured personal-data patterns can be masked before external analysis while the institution retains the original under its own controls.

3. Evidence collection

The platform gathers source matches and available integrity indicators, including citations, translated-text signals and mixed-script substitutions.

4. Structured review

Reviewers see locations and explanations, can record exclusions and notes, and apply the institution’s policy.

5. Fair process

Where appropriate, the institution can request a student response and a second review before closing a case.

6. Reporting and retention

Stable reports, audit events and retention controls support quality assurance and accountable decision-making.

Automated signals do not make the decision.

Similarity, AI-writing and manipulation indicators can be incomplete or wrong. Unplag.bg presents them as review evidence and keeps the human decision explicit.

Discuss your institution’s workflow.

We will show the product, answer technical and governance questions, and prepare a proposal around your actual needs.

Request a demo