Energy performance forecast is significant in optimizing energy allocation and integrating renewable energy sources. In this study, the authors employ Extreme Gradient Boosting (XGBOOST) machine learning combined with Explainable Artificial Intelligence (XAI) feature importance for energy prediction. We picked XGBoost because of its model's speed at handling large amounts of data, as well as its ability to deal with intricate, non-linear relationships. It was benchmarked using open-access datasets which proved its scalability and processing speed appropriate for big data. Explicability advances the importance of features' predicative role through XAI, thus increasing model transparency. This transparency improves stakeholder trust by showing primary factors of electricity production trends. Moreover, the model demonstrates the impact of external changes such as policies and technological advancements on power generation. To meet the emission reduction and net-zero carbon objectives, a changeover from fossil fuels, while exponentially increasing the renewable energy sources used, is vital. With prospective Machine Learning in the picture, operational expenses of energy projects get sliced due to precision in predicting energy output, especially with XGBoost onboard. The focus of this study is to develop defensible, yet detailed guidance for better energy management decisions. This research projected a new paradigm for addressing the efficiency and clarity of energy system forecasts using XGBoost coupled with XAI feature importance attaining (MAE) Mean Absolute Error of 3.83, (RMSE) Root Mean Square Error of 4.66 and (MAPE) Mean Absolute Percentage Error of 62.23%. © 2025 IEEE.