The purpose of this research is to develop an innovative technique to predict energy consumption with the help of machine learning using the household appliances as dataset. The project investigates certain Machine Learning practices, for example, the usage of household energy consumption’s time series prediction through LSTM and CNN based deep learning models. The methodology also covers steps such as Data Cleaning, Data Feature Engineering, Model Training, and Model Evaluation based on MAE, R2, and other performance indicators. The results of using deep learning models have shown an enhanced ability to accurately predict energy consumption patterns with greater explainability attributed through the usage of SHAP and LIME. This study shows how machine learning can help in predicting energy usage which can reduce costs and improve sustainability along with optimizing power distribution.