Responsible AI
Automation must strengthen human accountability, not obscure it.
Human authority
AI-supported signals may help authorized teams organize review, but they do not replace professional judgment or make final decisions about students, families, eligibility, discipline, or services.
Purpose limitation
AI use should be tied to a defined student presence task, use the minimum necessary data, and remain subject to tenant policy, legal review, and role-based access.
Explainability and challenge
Operational users should be able to understand the source information behind a signal, correct inaccurate records, document contextual factors, and override automated recommendations.
Outcome claims
Restora distinguishes workflow completion and observed outcomes from proven causation. AI-generated summaries or analyses must not be treated as validated impact evidence without appropriate evaluation.
Where AI may assist
AI may assist authorized staff by organizing case review, summarizing record content, drafting outreach language for staff review, and surfacing patterns across records. Assistance is always prepared for review by an authorized professional.
Where AI may never make the final decision
AI does not decide student eligibility, discipline, service authorization, case closure, family contact, or fund approval. These decisions remain with authorized staff under district policy.
Data and model handling
AI runs on authorized, tenant-scoped information using the platform’s language model integration. The minimum necessary data is used for a defined task. Restora does not represent that customer records are used to train third-party models.
Human override and review
Staff can review the source records behind AI assistance, correct inaccurate information, add context, and override any recommendation. Overrides and context are preserved in the record.