Fall is the second leading cause of unintentional mortality and injury across the globe. It is reported that the fall-related deaths toll 38,742 in the United States in the year 2021. Falls shall lead to fractures, head injuries, immobility etc. however cameras are avoided in privacy-aware environments which may lead to unnoticed conditions of fall. As wearable based devices are generally unsuitable for moist environments such as restrooms, the proposed system is designed to be wall-mounted which enables the elderly activity sensing without any physical contact. Moreover, camera-based fall detection solutions are not ideal due to concerns regarding user privacy. In this work, an Internet of Things (IoT) enabled fall detection system has been proposed for elderly privacy-aware environments namely restrooms, bedrooms etc. Furthermore, a non-contact type indoor fall detection sensor is interfaced to ESP32 IoT microcontroller which senses various elderly postures such as standing, sitting, absence and fall. The entire process is coded using an embedded C program and deployed to the IoT controller. Results demonstrate that the proposed system is capable of detecting elderly activity in a privacy-aware environment with high accuracy, without relying on intrusive camera-based monitoring. Also, the proposed system senses elderly in both static and dynamic states which outperforms the conventional Passive Infrared (PIR) sensor-based system. Future enhancements, including acquisition and transmission of elderly physiological conditions namely heart beat rate and respiration rate to the caretaker reduces the delaying medical response.