EduSphere AI: A Context-Aware and Policy-Constrained Approval Workflow Automation Framework for Higher Education Institutions

Approval paperwork in universities is slow for a simple reason: nobody agrees on where a document should go next. An examination schedule, a timetable change and a departmental notice all need sign-off, but from different officials, and the route is usually decided by whoever is holding the file. Fixed digital workflows add tracking but send every document down one path, which burdens simple documents and can bypass authorities on complex ones. This paper presents EduSphere AI, a framework that reads a document’s context (type, creator role, department, scope and impact), proposes an approval route, and checks that route against encoded institutional policy before any task is assigned. The system only recommends; authorised officials still make every decision. We formulate route selection as minimisation under policy constraints, describe the validation procedure and a policy-record schema, and design an evaluation against manual and fixed-route baselines that reports efficiency and compliance together. An analysis of five expected routes shows that a single fixed route would add unnecessary approvals in four document types and skip a required authority in one, while context-aware routes need one-third fewer approval steps (10 versus 15). A partial prototype covers role-based login and rule-based routing; the experiment is still pending.