Designing an Adaptable Mobile GIS with Conversational Assistant for Navigating Dynamic Campus Environments

University campuses are dynamic environments where frequent infrastructural changes create significant wayfinding challenges for students, staff, and visitors. Existing navigation tools often lack the localized data and adaptability required for campus-specific contexts. A preliminary survey at Gombe State University revealed that over 70% of students reported difficulties locating buildings due to recent renaming. This study presents GSU Connect, a mobile-based navigation system and case study in designing adaptable campus wayfinding solutions. Built with Flutter, Supabase, and OpenStreetMap, GSU Connect dynamically incorporates renamed buildings and new facilities to ensure accurate, real-time navigation. We further introduce GSU Assistant, a lightweight conversational interface powered by a rule-based model that enables users to retrieve spatial information through natural language queries. Unlike resource-intensive AI systems, the assistant requires no extensive training or computational overhead, making it accessible and efficient for campus-scale applications. A pilot deployment demonstrated high usability and efficiency, with task success rates ranging from 85% to 99% and average response times under three seconds. Findings confirm that a lightweight, open-source architecture can deliver scalable, cost-effective, and replicable navigation solutions for higher education. This work contributes both a functional prototype and a framework for extending such solutions to other institutions.