Falls among adults over 65 affects more than one in four each year, yet fewer than half ever report them, which creates life-endangering blind spots in senior healthcare. RiskWatch bridges those gaps with a layered detection and notification system: a wrist‐worn device running a TensorFlow Lite model on a Raspberry Pi 5 analyzes IMU and heart-rate data in real time, then broadcasts lightweight BLE advertisements to Raspberry Pi 3 Base Stations equipped with C920e cameras for vision-based confirmation. All events feed into Firebase for secure storage and an Android app interface, while Alexa skill delivers instant voice alerts on Echo devices. This combination of redundant fall detection methods, home-wide coverage, cloud archiving, and multi-channel notifications cuts false positives, extends coverage beyond phone range, and keeps caregivers informed the moment a fall occurs.