For institutions
Comprehension validation infrastructure for education.
Defensible evidence of learner understanding, governed by your policies, reviewed by your staff, retained on your terms.
The case for Vyvir
Evidence, not inference.
Detection tools tell you a document looks suspicious. They do not tell you whether the learner understands the subject, and they place your staff in the position of defending an inference.
Vyvir takes a different route. It asks the learner to explain, connect, justify and demonstrate, then hands your educators a structured record of what was shown.
That evidence is easier to review, easier to moderate and considerably easier to defend.
Governance built in
Your policies, not our assumptions.
Policy-driven
Assessment rules, attempt limits, integrity settings, interruption handling, evidence expectations and retention are all configured by you.
Access arrangements & reasonable adjustments
Approved access arrangements—including extra time, rest breaks, adjusted pacing and assistive technology—can be configured in line with institutional policy while preserving assessment requirements.
Policy versioning
The policy in force when an assessment begins is the policy that governs it. Later changes do not rewrite historical sessions.
Full audit trail
Assessments created and amended, questions approved, learners assigned, sessions started, interruptions, resumptions, outcomes and amendments. All recorded.
Moderation and quality assurance
Give internal quality assurers and moderators access to the evidence they need to review, without opening your entire institution to them.
Roles and delegation
Institution-level administration, optional departmental delegation, educators, assessors, moderators and learners, each with the permissions appropriate to them.
Integrity without surveillance
Vyvir is not a proctoring platform, and we will not sell it as one.
Where your policy requires it, a formal assessment can capture a small number of still images at unpredictable intervals to evidence that the authorised candidate remained present. Learners are told in advance. The number, the frequency and whether the feature is used at all are yours to set, and the feature can be switched off entirely.
Images are never used for emotion detection, personality analysis, biometric profiling or anything unrelated to the assessment. No single photograph, dropped connection or browser event establishes misconduct. Integrity evidence is flagged for authorised human review, and your existing academic integrity procedure takes it from there.
Integrity evidence supports educator review. It does not replace academic judgement.
Privacy by design
Only what is needed crosses the line.
Where it is technically and operationally feasible, the original learner submission stays inside your environment or the learner’s. Only the minimum information needed to conduct meaningful questioning crosses the boundary.
Your assessment data belongs to your institution. Where a learner also studies elsewhere, the two institutions share nothing.
Reliability
Reliability your registry can live with.
Attempts are explicitly defined, retakes are authorised rather than assumed, and interrupted sessions are recoverable. Every attempt carries its own provenance, so a retake is never confused with a first sitting.
Adoption at your pace
We are comfortable starting small.
Most institutions start with a single department or programme area, prove the model on a real cohort, then widen.
Uphold academic integrity at scale.