Business Process Management (BPM) is an approach to analyze and improve main activities of a company continuously. It seeks consistent results aligned with the strategic objectives. There are several approaches to the application of BPM, which focus on specific aspects of the company, not meeting all their needs. The Strategy, Indicators and Operations Model (MEIO), developed by Müller (2003), compiles the fundamental points of each isolated approach, creating a single model. The objective of this work is to provide a step-by-step application of the BPM aspects of MEIO in a practical case: a reference company that provides health services. Also, it provides a framework for organizing and connecting the various components of a product or service to its value chain. Through MEIO, a general analyze of the Company under study was made and the process “payment of invoices from providers” was detailed. Improvements were suggested based on a deeper investigation of the activities involved. The results include: (i) creation of rules engine to validate the procedures to be launched according to coverage, shortage, and contract; (ii) receiving procedure and cost audits separately and; (iii) digitalization and automatization of repetitive and manual activities.

The research aimed to investigate the stages of a Machine Learning model process creation in order to predict the indicator over the number of medical appointments per day done in the area of ​​supplementary health in the region of Porto Alegre / RS - Brazil and to propose a metric for anomalies detection. Literature review and applied case study was used as a methodology in this paper, besides was used the statistical software called R, in order to prepare the data and create the model. The stages of the case study was: database extraction, division of the base in training and testing, creation of functions and feature engineering, variables selection and correlation analysis, choice of the algorithms with cross-validation and tuning, training of models, application of the models in the test data, selection of the best model and proposal of the metric for anomalies detection. At the end of these stages, it was possible to select the best model in terms of MAE (Mean Absolute Error), the Random Forest, which was the algorithm with better performance when compared to Linear Regression and Neural Network. It also makes possible to identified nine anomaly points and thirty-eight warning points using the standard deviation metric. It was concluded, through the proposed methodology and the results obtained, that the steps of feature engineering and variables selection were essential for the creation and selection of the model, in addition, the proposed metric achieved the objective of generates alerts in the indicator, showing cases with possible problems or opportunities.