Geotechnical engineering is at the forefront of modern construction projects, as the stability and reliability of structures depend on a deep understanding of the underlying soil and geological conditions. Recent advancements in machine learning have revolutionized this field, offering a wealth of opportunities to enhance the efficiency, accuracy, and predictive capabilities of geotechnical engineering. There is much available literature that explores the diverse applications of machine learning in geotechnical engineering, including soil classification and prediction, site suitability assessment, site characterization, soil properties and behaviour. For this purpose, machine learning techniques like Linear Regression (LR) Analysis, Artificial Neural Network (ANN), Support Vector Machine (SVM), K-NN regression (KNN), Random Forest (RF) and M5 Tree (M5P) have been used. This paper delves into the specific methods and tools for the classification and characterization of soils using ML and highlights the potential recognition of the role of ML in the characterization of heterogeneous soil. By harnessing the power of machine learning algorithms, geotechnical engineers can unlock new insights from vast datasets, improve site characterization, and make informed decisions that lead to safer and more cost-effective construction projects. The review presented here will aid in the use of this tool for varied applications, highlighting real-world examples and the potential impact of these innovations on the future of geotechnical engineering.