Upgrading decision support systems with Cloud-based environments and machine learning

• Autor(a) principal: Almeida, Gonçalo Vagos Morais Callé de
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/119345
• topic: Business Intelligence; Microsoft; SQL Server; SQL Server; PostgreSQL; Cloud; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

Business Intelligence (BI) is a process for analyzing raw data and displaying it in order to make it easier for business users to take the right decision at the right time. Inthe market we can find several BI platforms. One commonly used BI solution is calledMicroStrategy, which allows users to build and display reports.Machine Learning (ML) is a process of using algorithms to search for patterns in data which are used to predict and/or classify other data.In recent years, these two fields have been integrated into one another in order to try to complement the prediction side of BI to enable higher quality results for the client.The consulting company (CC) where I have worked on has several solutions related to Data & Analytics built on top of Micro Strategy. Those solutions were all demonstrable in a server installed on-premises. This server was also utilized to build proofs of concept(PoC) to be used as demos for other potential clients. CC also develops new PoCs for clients from the ground up, with the objective of show casing what is possible to display to the client in order to optimize business management.CC was using a local, out of date server to demo the PoCs to clients, which suffered from stability and reliability issues. To address these issues, the server has been migrated and set up in a cloud based solution using a Microsoft Azure-based Virtual Machine,where it now performs similar functions compared to its previous iteration. This move has made the server more reliable, as well as made developing new solutions easier forthe team and enabled a new kind of service (Analytics as a Service).My work at CC was focused on one main task: Migration of the demo server for CCsolutions (which included PoCs for testing purposes, one of which is a machine learning model to predict wind turbine failures). The migration was successful as previously stated and the prediction models, albeit with mostly negative results, demonstrated successfully the development of large PoCs.