The goal of this project is to determine if it is possible to predict bankruptcy in companies two years before it happens. We created a model with financial and non-financial variables in order to evaluate its prediction capacity wih machine learning algorithms. Before starting the project, it has been necessary previous investigation about the state of the art and different studies done. After this obtaining the correct data has played a key role , which we have obtained from SABI database. Once the data has been gathered and the variables selected, data treatment was necessary so that we could apply the algorithms correctly. Finally, the results obtained demonstrate that it is possible to predict insolvency with a precisiion of around 97%.