On the Application of SCRUM in Data Science Projects

The emerging discipline of Data Science poses several challenges for teams conducting projects in the field as notably, the majority of Data Science teams fail to deliver the expected outcomes. To improve the results, researchers tried to adapt agile project methodologies like Scrum for Data Science projects. Scrum in particular is often implemented due its success in software engineering. However, the basic Scrum framework has proven itself to be too strict for Data Science, due to frequent unpredictabilities of Data Science tasks. Consequently, adaptions were made to traditional Scrum to make it more suitable for the new challenges. This article discusses further adaptations and suggests that Scrum in itself is usable in Data Science, however, additional adaptations of the core concepts need to be envisioned.