Nowadays, the application of new technologies in daily life is becoming more and more popular with the ultimate goal of improving people’s life experience. As a result, the number of electrical and technological devices in buildings is increasing, leading to the problem of managing and monitoring the level of electricity consumption. Thanks to the capabilities of IoT devices such as smart plugs, data on electricity consumption can be easily collected. However, this amount of data can be extremely big and needs to be processed in accordance with the purpose of use, sometimes immediately. One of the problems that is attracting attention is how to take advantage of current energy consumption data to detect abnormalities in electricity use to help power companies develop appropriate energy management plans. Thus, the paper has designed and implemented a lightweight system for real-time analyzing energy consumption and detecting anomalies in smart homes. The component of anomaly detection is built using statistical methods by comparing the energy consumption of one house with that of all other houses, allowing for quick and effective anomaly detection. The experimental results confirm the potential application of the system in smart energy solutions.