As energy grids become more complex, the need for effective monitoring the grid and security measures for utility services is growing. This research focuses on anomaly and theft detection in smart energy grids using data analytics and machine learning techniques applied on the real-world Transmission and Distribution (T&D) data by Supervisory Control and Data Acquisition (SCADA) data obtained from Madhya Gujarat Vij Company Limited (MGVCL). The methodology proposes an analytical unit for data preprocessing, anomaly detection with single-class classifiers, and forecasting key metrics. Anomaly detection identifies potential mishaps, which could signal faults or illegal usage. Machine learning models are used to predict future energy patterns, helping to improve grid’s robustness. Additionally, a dashboard based on Microsoft Power BI is proposed for real-time visualization of observed patterns. The study also discusses the implementation of smart meters to collect and monitor data in real time. This approach aims to enhance the security and efficiency of energy grids by providing timely detection of potential issues.