The primary objective of the research project is to employ machine learning algorithms to conduct studies and identify instances of phished URLs that might direct people to fraudulent websites. The Kaggle repository, which contains more than 11,000 URLs, is where the authors received a phished dataset for this application. Examples of both genuine and phished URL links can be found in the collection. Also, the dataset contains 31 features that must be obtained using feature engineering stages and methodologies. Nevertheless, this dataset is also available as a csv file and has been further pre-processed to remove redundant and pointless data. This is followed by the feature extraction process, which extracts URL properties including domain-based, content-based, and address-based attributes. The implementation of PyCaret follows, with each line of code being in charge of the entire execution. Nonetheless, the testing at this level consists of three parts. In order to create accuracy, the initial stage of PyCaret's implementation includes running 14 built-in algorithms. The top three accuracy-generating algorithms are combined to build a stacking model in the last stage of the system model's implementation, which is divided into two stages. The second stage of the system model implementation entails taking random forest into account. In the conclusion, the accuracy of each algorithm is assessed together with its performance. After comparison, the technique with the highest generating accuracy is considered to be the optimised model.