The ShatterTattler is an Internet of Things (IoT) home security device that detects glass breaks and notifies the end user of potential break-ins and security risks. This project aims to create a smart home device that detects glass breakage with high precision using AI and Digital Signal Processing (DSP). The system triggers an alarm, visual alerts, and sends real-time notifications to the end user, all while being wirelessly connected and battery-powered. All electronics are housed and mounted in a custom 3D-printed case to ensure the system remains stable and robust. For our project, our team implemented three key technologies. The first key technology used is the Jetson Orin Nano. This development board is used in the design for its computing power. The Jetson Orin Nano enabled our team to implement an event detection AI model. Second, the ESP32-S3 microcontroller is used. This microcontroller is used for its compactness, low power consumption, and its Bluetooth communication capabilities. Finally, LiPo batteries are incorporated into the design for their high energy density, compactness, and modularity. The LiPo batteries are used to regulate the system's power supply at 3.3V. Our notable accomplishments include successfully implementing a detection and classification machine learning (ML) model that accurately classifies audio signals from the ambient environment and alerts the end user of dangers. Our team appreciates the flexibility in our design. This allows for this design to be adapted for other use cases. Beyond our successes and challenges, we acknowledge room for improvement and growth as designers. For greater portability and to match similar on-the-market products, the enclosure size should be reduced. Next, the team recommends that the product be adapted to other environments, which would require data that can classify various event types. Next, to enhance user experience, our team recommends creating a web interface for system configuration and monitoring. Lastly, our team recommends strengthening the ML model by integrating adaptive learning to increase event detection.