Guiding software companies into the future

Digitalisation is fundamentally changing software-intensive companies, especially those developing embedded systems, by focusing on models that leverage data and AI through machine learning and deep learning.

... many companies are still struggling to transform their models from prototypes to fully functional operational systems that can be used in the market.

Meenu Mary John

“Despite these advances, many companies are still struggling to transform their models from prototypes to fully functional operational systems that can be used in the market,” says Meenu Mary John, a doctoral student at the Department of Computer Science and Media Technology.

Her thesis, Towards Continuous Development of MLOps Practices, concerns developing systematic and structured frameworks for deploying and commissioning so-called ML/DL models (where ML/DL stands for Machine Learning and Deep Learning respectively) within these software-intensive companies.

She has investigated the challenges experienced, and the activities performed by companies during the implementation and use of these processes. To achieve that, she has used a combination of different empirical research methods, such as case studies, action research, and literature reviews.

“My findings can hopefully make even the less experienced system developers and software engineers become better at the design and development of ML/DL models. The frameworks are certainly not just for the data scientists and ML experts,” she says.

The research has been conducted within the framework of the Software Center, which is a collaboration between fourteen companies (including Ericsson, Volvo, Tetra Pak, and Bosch) and five Swedish universities – which includes Malmö University – with the mission of improving the digitalisation capabilities of the European software-intensive industry.