AI changed what students can generate. Now we must ask, how can assessment reliably demonstrate evidence of learning?
Meandrix is a higher-education assessment platform built for a calm, defensible premise: in an AI-enabled world, assessment must return to evidence of reasoning in action, not just answers at a point in time.
The issue isn’t AI. It’s what counts as evidence.
When the answers to knowledge recall tasks become easy to generate, the reliability of standard education assessments drop. A common response is heavier surveillance — but surveillance-first approaches create avoidable cost, accessibility burdens, and trust damage.
What institutions need is a scalable way to capture applied judgement and produce artefacts that support moderation, quality assurance, and accreditation narratives.
Our position: “AI broke assessment → evidence matters”
Meandrix is built around assessments where learners must decide in context. Those decisions carry weight, shape outcomes, and generate a traceable record of performance.
Context first — prompt judgement rather than memorisation.
Decision-weighted movement — outcomes reflect choices and trade-offs.
Defensible artefacts — evidence that supports review, feedback, and assurance.
Design principles
This is an institution-first product. The aim is not hype, but a system that fits governance, scrutiny, and academic judgement.
Evidence over theatre — prioritise signals that hold up to review.
Integrity by design — disincentivise misconduct without costly or intrusive surveillance.
Transparent logic — interpretable scoring and pathways support moderation.
Feedback that moves learning — decision-linked guidance, not generic commentary.
Governance-ready — built for auditability, review, and defensible adoption.
Origin and intent
This platform was shaped inside the practical realities of higher education: scale, policy, scrutiny, and the need for defensible outcomes.
Meandrix was created by an educator working with high-stakes assessment and large cohorts, where consistency, moderation, and evidence quality matter as much as learner experience. The aim is not to replace academic judgement — it is to provide infrastructure that makes judgement easier to exercise and easier to defend.
Integrity, proportionately
Meandrix is designed for integrity without creating a surveillance product. Where additional deterrence is needed, it should be policy-aligned, accessible, and proportionate.
Optional integrity controls (e.g., Sentinel) are intended to discourage misconduct while preserving learner trust and institutional credibility. Operational details are best discussed in an institutional demo to align with local policy, device constraints, and equity considerations.
Want to evaluate Meandrix for your institution?
We’ll walk through evidence outputs, moderation workflows, and integrity settings—calmly and concretely.