<![CDATA[Background: The advancement of digital technology has transformed government administrative processes, demanding more transparent and accurate systems for managing employee attendance. Specific background: The Tomohon Land Office still relies on manual attendance methods that often result in inefficiency, fraud potential, and limited verification capability. Knowledge gap: Current attendance systems lack integration between geographical validation and automatic identity verification. Aims: This research aims to develop an attendance system based on Global Positioning System and Deep Learning algorithms for reliable facial recognition and real-time location validation. Results: Using the Waterfall development model, the system integrates geographical coordinates and facial data, achieving 95–98 percent recognition accuracy and under two seconds of response time. Novelty: The system introduces a combined method of multi-angle facial enrolment and precise location verification through Haversine calculation. Implications: This innovation promotes transparency, accuracy, and efficiency in employee attendance management and supports digital transformation in public institutions. Highlights Integration of Global Positioning System and Deep Learning enables precise attendance validation. Multi-angle facial enrolment ensures stable and reliable recognition results. Supports digital transformation for transparent and efficient attendance management. Keywords Global Positioning System, Deep Learning, Facial Recognition, Employee Attendance System, Digital Transformation]]>
