Authors:
Christian Haertel
;
Christian Daase
;
Daniel Staegemann
;
Abdulrahman Nahhas
;
Matthias Pohl
and
Klaus Turowski
Affiliation:
Magdeburg Research and Competence Cluster VLBA, Otto-von-Guericke University Magdeburg, Magdeburg, Germany
Keyword(s):
Data Science, Project Management, Cloud Computing, MLOps, Automation.
Abstract:
The significant increase in the amount of generated data provides potential for organizations to improve performance. Accordingly, Data Science (DS), which encompasses the methods to extract knowledge from data, has increased in popularity. Nevertheless, enterprises often fail to reap the benefits from data as they suffer from high failure rates in the conducted DS projects. Literature suggests that the main reason for the lack of success is shortcomings in the current pool of DS project management methodologies. Hence, new procedures for DS are required. Consequently, in this paper, the outline for a model for DS project standardization and automation is discussed. Following a summary of DS project challenges and success factors, the concept, which will incorporate MLOps and cloud technologies, and its individual components to address these issues are described on a high level. Therefore, the foundation for further research endeavors in this area is presented.