MLOps: Machine Learning Development and Operations

Thesis (BA/MA)

Advisor(s): Michael Weber (mic.weber@tum.de)

Context

Many of today’s systems labelled “Artificial Intelligence” rely on machine learning (ML) techniques. ML enables machines to learn and improve from experience, such as historical datasets (Jordan & Mitchell, 2015). While ML offers manifold opportunities for business and society, many organizations are stuck in the experimental phase and struggle with ML development and operation (Benbya et al. 2020). Research suggests that today’s ML systems differ from conventional software systems through their increasing 1) autonomy, 2) learning ability, and 3) incomprehensiveness to certain audiences (Berente et al., 2021). These unique characteristics require organizations to take different approaches to software development and operation. For example, ML development and operation were found to require intimate collaboration between domain and ML experts (van den Broek et al., 2021), continuous feedback loops to maintain performance (Gronsund & Aanestad, 2020), and specific coping strategies for opaque “blackbox” ML systems (Asatiani et al., 2021).

However, our understanding of ML’s impact on software development and operation is still at an early stage and many research issues remain open (cf. Berente et al. 2021). Therefore, the goal of this thesis is to contribute to our understanding of emergent practices, success factors, and challenges in ML development and operation. A particular focus shall be laid on understanding how ML’s unique characteristics potentially require novel or different approaches to conventional software development and operation. For this, a variety of research designs could be applied, including 1) systematic reviewing of literature, 2) case studies of real-world ML projects, 3) case studies of organizational ML initiatives, or 4) survey design and execution on project- or organizational-level. The exact scope of the thesis and its research design will be iteratively developed between you and your advisor.   

Task(s)

  • Review literature in the respective field
  • Collect and analyze empirical data (e.g., interviews, archives, surveys)
  • Assess novelty of findings regarding ML’s unique characteristics
  • Discuss implications for Information Systems research

Requirements

  • High degree of autonomy and individual responsibility
  • Good communicative skills
  • Basic understanding of machine learning is beneficial

Further Information

The thesis can be written in English or German. The topic can also be adapted to your interests. If you have further questions, please do not hesitate to contact me directly. Please send your application including our application form, a current transcript of records, and your CV to mic.weber@tum.de. Please note that we can only consider applications with complete documents. 

Literature

Asatiani, A., Malo, P., Nagbøl, P. R., Penttinen, E., Rinta-Kahila, T., & Salovaara, A. (2021). Sociotechnical Envelopment of Artificial Intelligence: An Approach to Organizational Deployment of Inscrutable Artificial Intelligence Systems. Journal of the Association for Information Systems, 22(2), 325-352.

Benbya, H., Davenport, T. H., & Pachidi, S. (2020). Artificial Intelligence in Organizations: Current State and Future Opportunities. MIS Quarterly Executive, 19(4), ix-xxi.

Berente, N., Gu, B., Recker, J., & Santhanam, R. (2021). Managing Artificial Intelligence. Mis Quarterly, 45(3), 1433-1450.

Grønsund, T., & Aanestad, M. (2020). Augmenting the algorithm: Emerging human-in-the-loop work configurations. The Journal of Strategic Information Systems, 29(2), 101614.

Jordan, M. I., & Mitchell, T. M. (2015). Machine learning: Trends, perspectives, and prospects. Science, 349(6245), 255-260.

van den Broek, E., Sergeeva, A., & Huysman, M. (2021). When the Machine Meets the Expert: An Ethnography of Developing AI for Hiring. Mis Quarterly, 45(3), 1557-1580.