Placeholder case study. Replace the client details, figures and narrative below with the approved version before publishing.
The situation
Examination results were assembled from spreadsheets submitted by affiliated colleges, consolidated centrally by hand, and published roughly three months after the final paper. Errors surfaced after publication, and each correction triggered a manual revaluation cycle.
What we built
A single platform covering examination form submission, centre allocation, evaluation entry with double-blind verification, moderation, and result computation against the university’s own ordinance rules.
The ordinance rules — grace marks, back-paper eligibility, moderation bands — were implemented as versioned, dated configuration rather than code, so the examination controller can adjust them after an academic council decision without a software release.
Outcome
Result publication moved from roughly three months to three weeks. Errors caught at the verification stage rather than after publication. Every mark carries an audit trail back to the evaluator who entered it.