Over the last decades, the problem of detecting leaks in pipelines with the help of software has been much discussed. Leak detection in pipelines is a critical issue, with significant economic and environmental implications. Timely detection can prevent water loss and mitigate potential disasters. Traditional methods, such as leakage control policies, detection systems in control centers, and field brigades, are effective but do not offer real-time detection. This paper presents a machine learning-based approach for detecting pipeline leaks using process data from the pipeline system. Two machine learning techniques, Decision Trees, and Support Vector Machines (SVM) were implemented and compared. The Decision Tree model achieved an accuracy of 88%, while the SVM model demonstrated an f1_score of 0.7666 and a Jaccard Index of 0.8397. These results show that machine learning can improve the speed and accuracy of leak detection by leveraging real-time data, offering a more efficient and scalable solution for pipeline monitoring.