ST IT Cloud Success Cases
Education
Technology
Machine Learning / Cloud / AWS / SageMaker / AWS DMS / AWS Glue / Athena / AWS QuickSight
Goal
The objective of this project was to detect the dropout probability of students, through the analysis of the various data available from students, application of machine learning model training so that, in the end, each student in the base has a dropout probability score associated with their profile. The data used for the development of this project were made available by the educational institution, such as: financial data, data on the student's presence (entrance and exit times) and student history at the institution.
Education
Technology
Machine Learning / Cloud / AWS / SageMaker / AWS Glue / Step Functions / Neptune
Goal
The objective of this project was the development of a course recommendation engine, through the analysis of data from the educational institution's student database, tables of cities, table of municipalities, IBGE table, table of knowledge area, comparing Area x Course and Completion year table. After aggregating these data, Neptune was used to facilitate the creation and execution of applications with graphs. The data used for the development of this project were provided by the educational institution, with the support of public data.
Energy
Technology
Machine Learning / Cloud / AWS / SageMaker / AWS DMS / AWS Glue / Athena / AWS QuickSight
Goal
Develop Machine Learning models to identify new consumption variation profiles, categorize consumers, automate risk analysis and expand your product portfolio. Public data from the CCEE (Electric Energy Trading Chamber) were used to classify the consumer consumption variation profile via CNPJ and subsequently apply metrics (provided by the customer) to establish premiums, discounts or to apply any marketing actions with these customers.
Insurance and Education
Technology
Machine Learning / Cloud / AWS / SageMaker / Textract / Rekognition / Comprehend
Goal
Develop Machine Learning models for:
• Image treatment – The image treatment phase will be responsible for classifying, typing, validating the legibility of the document, correcting angle and perspective, improving image quality (adjustments to brightness, contrast and resolution), conversion to specified format ( PDF/A), automatically and provide only the information required by the system, along with the document score.
• Image manipulation – perform all image validation processes, OCR and training of Machine Learning models.
English 
Portuguese 
Spanish