The rapid growth of electronic communication within higher education institutions has significantly increased the volume of emails processed by administrative personnel. Consequently, critical institutional communications are often overlooked among routine messages, promotional content, and spam, resulting in communication delays, reduced productivity, and administrative inefficiencies. This study presents the design and implementation of an Artificial Intelligence (AI)-based email filtering and prioritization system tailored for polytechnic administrative environments. The proposed framework integrates Natural Language Processing (NLP) and supervised machine learning algorithms to classify incoming emails into Urgent, Actionable, Routine, Promotional, and Spam categories. Multiple classification algorithms, including Naïve Bayes, Support Vector Machine (SVM), Random Forest, and DistilBERT, were evaluated to determine optimal performance for institutional communication datasets. The system incorporates contextual prioritization mechanisms, role-based access controls, audit logging, and a web-based dashboard for real-time communication management. Findings indicate that intelligent email prioritization can significantly improve communication responsiveness, reduce email overload, and support digital transformation initiatives in tertiary institutions. The study contributes a scalable framework suitable for deployment across Nigerian polytechnics and other public-sector organizations.