Cloud Business Intelligence

In this thesis, I design a business intelligence solution in a cloud for a medium-sized company. As a suitable approach to creating a data warehouse, I chose the incremental building of data markets connected by conformed dimensions. I was inspired by the implementation methodology by R. Kimball, but I modified it for the needs of a smallerscale project built in the cloud. First, I determined the needs and requirements for the BI solution through interviews and determined the availability and quality of data in the source systems. Furthermore, I designed the architecture of the solution using the ELT approach and selected the appropriate products: Cloud Data Fusion, Keboola, BigQuery, Dataform, Looker Studio, and Colab. Subsequently, I designed a dimensional model for the priority requirement. Next, I designed the structure of the BigQuery database, designed the data pumps, and developed them using the Dataform framework. The last steps were the report's design and implementation and the documentation creation. The result is a functional solution that meets the identified needs and the first implemented requirement from connecting data sources to the data marketplace and the resulting report in Looker Studio. The chosen architecture and used products can be used by many other companies or organizations, mainly thanks to good scalability and low input and operating costs.