The rise in theft incidents within institutional spaces has prompted the need for innovative security solutions. In response to this challenge, our research focuses on the development and implementation of a comprehensive theft prevention system through object and pose detection technologies. We employ cutting-edge techniques and models to safeguard institutional property and create a secure environment. For object detection, we leverage the powerful ”Segment Anything” model, which enables us to identify and track objects within the institutional space. This model provides us with a robust foundation for monitoring and safeguarding valuable items. In our pursuit of advanced object detection and classification, we explore the capabilities of multiple machine learning models, including Ridge, Logistic, Random Forest, and Gradient Boosting. These models enhance our ability to accurately classify objects and further strengthen our theft prevention strategies. Additionally, we utilize the state-of-the-art Media pipe Holistic model for real-time pose detection, enabling us to identify human poses and behaviors within the institutional space. This valuable insight adds an extra layer of security by recognizing suspicious activities and potential threats. Our research encompasses a holistic approach to security, integrating object and pose detection to ensure the highest level of theft prevention. By combining these technologies, we aim to significantly reduce theft incidents and enhance security within institutional spaces. As we continue to advance our research, we anticipate future challenges and complexities related to the integration of these technologies. This research sets the stage for ongoing exploration and innovation in the realm of institutional security, ultimately contributing to safer and more secure environments.