Currently, the application of a Machine Learning approach by researchers or companies in a project is still considered a process with a high level of difficulty. This is because it is necessary to know in some detail the algorithms and how they can solve the problem. In addition, it is important to find the technology that meets the needs of computational project, which support certain algorithms and enabling a good performance with large amounts of data. So, this paper, a work in progress, intends to introduce a framework to help and, in a way, guide researchers on a project of Machine Learning to choose the Use Cases, algorithms and technologies.