Internet of Things (IoT) term has been hyped for years, and takes so many thing's places in everywhere and makes many things to become easier and remotely controllable with smart automations. Human role in such areas are about to be vanished and sensors, actuators, gateways take over the workloads from human-being by generating the values which are desired to be measured, and transferring them within the network They also take decisions with some preset rules, artificial intelligence or machine learning methods on behalf of humans. Edge computing became a vital as there is a huge amount of requirement of low latency, extra resources, network restrictions, loose connections, real-time decisions, etc. In quick service restaurants, many waste management and service optimizations are human or paper-based which contains pre-calculated or pre-simulated values. In this paper, we propose an IoT architecture for quick service restaurants and describe various edge computing applications including processing the sensor values, extracting meaningful information, providing data integrity and more importantly learning the data patterns to present predictions, create alerts, or make some intelligent decisions to provide waste minimization and service optimization.