Reliability
Route success, provider health, queue completion, error rates, recovery time, email delivery, media availability, and incident resolution.
Pilot evidence
Evidence before endorsement.
Perfect Aptitude should be evaluated against the institution’s real roles, languages, content, devices, providers, governance requirements, and operating conditions. Published references belong here only after the speaker, organisation, wording, permission, date, scope, and commercial context are verified.
What a credible pilot should measure.
Usability by role
Task completion, navigation accuracy, mobile behaviour, accessibility findings, training needs, support burden, and approval clarity for each role.
Academic quality
Teacher review effort, language suitability, OCR and conversion accuracy, summary quality, question quality, curriculum fit, and publication corrections.
Governance
Permissions, institution boundaries, content rights, consent, retention, incident ownership, provider contracts, release evidence, and accountable approval.
Adoption and operations
Active use by role, completed workflows, unresolved blockers, support trends, operating cost, and whether the institution can sustain the configured process.
Decision criteria
Proceed, revise scope, extend validation, or stop—based on evidence agreed before the pilot rather than unsupported outcome claims.
Evidence package
A pilot should leave an auditable record.
Keep the scope, configuration, test accounts, scenarios, issue log, acceptance results, reviewer notes, unresolved risks, support ownership, and final decision together. This prevents a successful demonstration from being mistaken for a production-ready deployment.
- Signed scope and data-handling responsibilities
- Role-route and permission acceptance matrix
- Provider, queue, storage, email, and recovery evidence
- Accessibility and mobile test findings
- Teacher review and content-quality record
- Go, revise, extend, or stop decision
Discuss a controlled evaluation.
Define scope, acceptance criteria, responsibilities, data handling, provider dependencies, support, and exit conditions before starting.