The goal of this thesis is to investigate and define stakeholder engagement and involvement within the lifecycle of sensor-based MLOps. While MLOps principles streamline the deployment and maintenance of machine learning (ML) models, ensuring proper stakeholder involvement remains a challenge. Different stakeholders must collaborate effectively across various lifecycle stages to ensure model reliability, fairness, and operational success.
Thesis Type |
|
Status |
Running |
Presentation room |
Seminar room I5 6202 |
Supervisor(s) |
René Reiners |
Advisor(s) |
Michal Slupczynski |
Contact |
slupczynski@dbis.rwth-aachen.de |
MLOps extends DevOps principles to machine learning, ensuring reproducibility, scalability, and automation in ML pipelines. While existing research has focused on the technical aspects—such as model deployment, monitoring, and version control—less attention has been given to the human factors and stakeholder involvement in MLOps.
Prior research on stakeholder engagement in software engineering has identified techniques such as participatory design, agile feedback loops, and domain expert involvement. However, mapping these techniques to MLOps—particularly in sensor-based AI systems, where real-time data and domain expertise are crucial—remains an open research question.
By systematically analyzing existing literature, this thesis will provide insights into how stakeholder roles evolve in sensor-based MLOps and propose best practices for engagement and collaboration.
- Background knowledge in MLOps (basic understanding of ML pipelines and deployment).
- Experience with literature reviews and systematic mapping studies.
- Interest in stakeholder engagement, collaboration techniques, and human factors in AI.
- Basic knowledge of sensor-based AI applications (optional but beneficial).