This research presents a comprehensive exploration into the various methodologies and techniques related to urban energy modelling. A key focus is the potential enhancement of predictive accuracy in energy consumption through the integration of data-driven models with machine and deep learning approaches. An examination of multiple machine learning and deep learning algorithms, including Linear Regression, Decision Trees & Random Forests, Gradient Boost, and Deep Neural Networks, underscores the novelty of this study. The research capitalizes on advanced Python libraries such as TensorFlow, Scikit-Learn, and PyCaret, thus enriching the versatility and effectiveness of the conducted procedures. The implementation of Python, along with its libraries, facilitated the application of machine learning and deep learning techniques, while open-source data from the Dubai municipality provided diverse input data. Notably, the study introduces a novel approach that leverages the PyCaret library to optimize algorithm selection based on specific data, while also employing TensorFlow for deep neural networks. Ultimately, the research proposes an innovative approach that combines machine learning, deep learning, and automation to enhance the potential of urban energy modelling. The study points to more complex and accurate analyses, potentially redefining industry best practices.