With increasing demand for a reliable and auto-mated attendance system has increased the use of embedded sys-tems, facial recognition software, and the Internet of Things (IoT) in schools. In this paper we propose an intelligent attendance management system that is built upon Raspberry Pi, camera modules and performs real-time face detection, alignment, extrac-tion of features and recognition using optimized computer vision and deep learning techniques that can work on limited hardware. To adapt for the changes in light, device position and person expressions, it uses edge computing techniques and small models that enable fast computations (low latency). In educational setups, attendance records are maintained and accessed for real-time tracking, reporting, and analysis of data through web and desktop applications. Results from live trials are showing in-class performance, high recognition and low false acceptance rates. The suggested framework enhanced efficiency, accuracy, and reliability with minimal human involvement by offering a scalable, affordable, and fully automated alternative to traditional attendance systems.