This research article explores the application of machine learning algorithms to predict energy consumption in educational buildings, aiming to enhance energy efficiency and sustainability. The study focuses on three buildings at Down Town University: CLAS, NHAI, and Cronkite. The data collected from January 2020 to January 2023 is analyzed using three machine learning models: long short-term memory (LSTM), random forest (RF), and gradient boosting regressor (GBR). The models are evaluated using performance metrics such as root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The research begins with an introduction to the integration of AI in smart buildings, highlighting the importance of energy management. It reviews existing literature on energy consumption forecasting using various machine learning algorithms, including traditional statistical methods and deep learning approaches. The methodology section details the data analysis, model development, and evaluation processes. Data preprocessing involves handling missing values and standardizing the data, while feature selection identifies the most influential parameters. The results show that GBR outperforms both LSTM and RF in predicting energy consumption, with the lowest MAPE values of 4.045, 12.338, and 9.337 for CLAS, NHAI, and Cronkite, respectively. LSTM and RF also perform well, with lower MAE values for some buildings. The study concludes by discussing the implications of the findings and suggesting future research directions, including optimizing GBR models and applying transformer models for diurnal energy consumption prediction.This research article explores the application of machine learning algorithms to predict energy consumption in educational buildings, aiming to enhance energy efficiency and sustainability. The study focuses on three buildings at Down Town University: CLAS, NHAI, and Cronkite. The data collected from January 2020 to January 2023 is analyzed using three machine learning models: long short-term memory (LSTM), random forest (RF), and gradient boosting regressor (GBR). The models are evaluated using performance metrics such as root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The research begins with an introduction to the integration of AI in smart buildings, highlighting the importance of energy management. It reviews existing literature on energy consumption forecasting using various machine learning algorithms, including traditional statistical methods and deep learning approaches. The methodology section details the data analysis, model development, and evaluation processes. Data preprocessing involves handling missing values and standardizing the data, while feature selection identifies the most influential parameters. The results show that GBR outperforms both LSTM and RF in predicting energy consumption, with the lowest MAPE values of 4.045, 12.338, and 9.337 for CLAS, NHAI, and Cronkite, respectively. LSTM and RF also perform well, with lower MAE values for some buildings. The study concludes by discussing the implications of the findings and suggesting future research directions, including optimizing GBR models and applying transformer models for diurnal energy consumption prediction.