Business Management Execution On Data Streams

Abstract

The world of business management is largely populated with data warehouses as a single source of truth. However, in recent years a shift towards the data origin known as data streams have arisen. Corporater, which is one of the leading companies in delivering business management solutions, acknowledges this trend and wants to investigate their possibilities in supporting data streams. The data streams are most valuable when they are processed and analyzed by a stream processor. This is because the singular events by them selves are less meaningful than a collection of manipulated events. Hence, part of the investigation includes finding the most suitable stream processor for Corporater. Further, it also must be proven that it is possible for stream processors to integrate with Corporater's systems and values.

Through this thesis, we aim to provide this investigation for Corporater. We achieve this by evaluating and developing a prototype which abides Corporater's requirements and environment. Furthermore, the evaluation provides a general overview of the leading architectures and processors for data streams from a business perspective, which conclusively recommends a solution for Corporater. Additionally, the generality of this evaluation allows for beneficial value of other businesses in a similar situation.

Based on the recommendation from the evaluation, we create the prototype. This prototype is largely focused around one of Corporater's core concepts. This is the `Business-In-Control' concept, which involves having the business experts manipulate and visualize data without the need of a developer. Thus, we implement the prototype in a generalized way that enables configuration from a different environment. This generality enables the prototype to be compatible with any type of GUI or API that are able to send configuration to the prototype. We test the prototype by performing an experiment. The experiment proves to be successful in creating an end-to-end connection with the data streams and Corporater's software. Hence, also proving that it is possible for today's stream processors to integrate with Corporater's system and values.