An essential part of planning a project is to define a time horizon in which the activities and procedures will be carried out. This is is even more relevant when it comes to organizing and handling relevant public documentation for the proper functioning of government entities. Consequently, is necessary a tool that makes it possible to estimate the duration of document management contracts, which is based on statistical methods that guarantee a good level of certainty for the articulation of the different plans proposed by the public entity. In this project, various alternatives were explored in order to maximize the precision and accuracy of the forecasts, starting with traditional techniques such as regression models, more specifically with Gamma regression and the Generalized Additive Models, and exploring the field of artificial intelligence with methods such as Random Forest and Classification and Regression Trees, finding different solutions that can be applied in different areas.