The world of Big Data continues to grow and evolve day by day and with it the technologies capable of handling huge amounts of data. The problem is to choose the right technology and exploit it in a way that optimizes processes working with these data. The aim of this project is to study and utilize the potential of a NoSQL system such as MongoDB for storing, extracting and analyzing large quantities of data concerning a system of credit card fraud detection. To do this, we generated plausible data, saved it in MongoDB, distributed it in the most efficient way for us, run several queries and finally analyzed the performance of the system. The results showed that the MongoDB database, if used in the right way, can efficiently manage large amounts of data. Based on the results, it is clear that in the world of Big Data, if the technologies to manage it are not used and optimized properly, the performance of the system becomes unacceptable. Specifically, in MongoDB, the modeling of collections and documents proved to be fundamental.