The modern power grid is rapidly changing because of technological breakthroughs and increased integration of renewable energy sources. Accurate energy consumption forecasting, often called load forecasting, is critical to the stability and efficiency of electrical distribution networks. This complexity poses challenges in terms of maintenance, calibration, and cost, highlighting the need for simpler and more streamlined approaches in model development. As a result, there has been increased interest in using machine learning methods to improve the accuracy and adaptability of load forecasting models. This work presents a comprehensive approach to improve load prediction accuracy considering the load details of all the states of India, by employing Support Vector Machine (SVM) and Random Forest (RF) models. The model is formulated with load and wind direction as inputs. The dataset comprises information like load, temperature, surface pressure, precipitation, wind pressure, and wind direction gathered from the entire country “India” and is narrowed down to include only two inputs: load and wind direction. This focused dataset allows for a simplified model that captures the essential dynamics of energy consumption while minimizing complexity. By considering these two crucial inputs, the model aims to provide accurate predictions of load values nationwide. A comparative study of both the algorithms in terms of its performance is presented. The RF model demonstrates satisfactory performance, exhibiting reduced susceptibility to overfitting and achieving greater accuracy with fewer inputs compared to existing models.