This project proposes a tool for analyzing, identifying, and determining potential combinations of datasets, focusing on the discovery of primary and foreign keys between two datasets (enabling relational joins), or similarity between datasets (enabling union, intersection, and difference operations). The approach leverages data mining and machine learning techniques to automate the correlation process between tables. By doing so, the project aims to enable new forms of data analysis that were previously unattainable due to the lack of explicit relationships between datasets. The results are expected to enhance data integration and uncover insights in contexts where datasets appeared unrelated.