Global energy patterns significantly influence the achievement of Sustainable Development Goals (SDGs) by driving access to clean, affordable energy while reducing greenhouse gas emissions to combat climate change. Sustainable energy use is essential for fostering economic growth, reducing poverty, and improving health and well-being across the globe. In this study, we have developed a Machine Learning (ML)-driven framework with supervised, unsupervised, and ensemble ML strategy. We have clustered the countries based on their level of achieving SGD and predict “energy intensity level of primary energy,” “access to electricity (% of population),” and “access to clean fuels for cooking” using different supervised ML models and proposed ensemble model. Our proposed method uses blending technique and integrates Decision Tree, Random Forest, Ridge, CatBoost together, named B_DRRC model. Proposed B_DRRC shows better performance compared to existing models. Finally, we have predicted the trend of the variables up to 2030 that shows the significant improvement of global energy pattern in SDGs. Further works focus on different SGDs and related variables to find the more accurate influence of the energy pattern in SDGs.