Industry 4.0 has brought innovative principles to the entire world, especially for the manufacturing industry. The adaptation to a technological era showed limitations in the current processes, of which we can highlight the divergence between software and machinery technologies, cloud data processing, difficulty for the information to circulate within a manufacturing environment, so that it flows clearly and objectively, without ambiguity. These limitations end up generating errors between operations in the manufacturing process resulting in costs, customer dissatisfaction, low product quality, and reduced competitiveness. Thus, problems related to the semantic web, semantic interoperability, horizontal and vertical integration are responsible for such limitations in manufacturing processes. To resolve such restrictions and improve the final quality of the product, it is possible to apply Machine Learning techniques. Through the use of ensemble models of machine learning algorithm techniques, techniques with specific characteristics can be grouped, complementing each other, thus providing better prediction results during the manufacture of products, reducing costs, increasing the reliability and quality of the final product. In this way, it is expected to improve the final quality of the product and minimize the impacts that detract from the performance indicators, such as scrap, cost, rework, labor. This research will contribute scientifically to the creation of a system, which can be applied in different manufacturing production processes.