Concrete piles are among the most important elements of deep foundations, as they are used to transfer structural loads to deep soil layers capable of bearing them. Accurate estimation of pile bearing capacity is a fundamental issue in geotechnical design, given its direct impact on the safety and economic efficiency of structures. With the rapid development of artificial intelligence technologies, machine learning has become a promising tool for analyzing geotechnical data and predicting pile bearing capacity, relying on the results of field and laboratory tests and on previous data. How Do Machine Learning Techniques Work? Machine learning models are based on training algorithms using data from piles that have been previously constructed or tested, after which they learn the relationship between the properties of the soil and the pile and the load the pile can bear. The most prominent data used in building these models include: Soil properties, such as type, strength, and the values of the Standard Penetration Test (SPT) or Cone Penetration Test (CPT). Pile dimensions, such as diameter, length, and cross-section shape. The method of pile installation and its type. Pile load test results, which are used to train the model and verify the accuracy of its predictions. Common algorithms in this field include Random Forest, Support Vector Regression, and Artificial Neural Networks. The Importance of These Techniques in Geotechnical Engineering Machine learning models help uncover complex relationships between soil properties and pile performance. They may provide rapid estimates of bearing capacity, support the comparison of design alternatives, and reduce the time required for preliminary analysis. However, the reliability of the results depends on the quality and diversity of the data and how well it represents site conditions. Therefore, the model’s performance should be verified using independent data and compared with the results of field tests and approved geotechnical design methods. Automated prediction is not a substitute for engineering judgment or the requirements of approved codes. Conclusion The use of machine learning to predict pile bearing capacity represents an important step toward developing intelligent geotechnical design. By integrating field data with modern algorithms, the accuracy of estimates can be enhanced and engineering decision-making supported, contributing to the design of safer, more efficient, and more economical foundations.