Project information
Doctoral student
Tibo Bruneel
Supervisor
Morgan Ericsson
Assistant supervisor
Jonas Nordqvist, Diego Perez
Participating organizations
Linnaeus University, Softwerk
Financier
SKF, The Knowledge Foundation (Industriforskarskolan för Data Intensive Applications + (DIA+))
Timetable
September 2023 –September 2028
Subject
Computer and Information Science (Department of Computer Science and Media Technology, Faculty of Technology)
Research group
Data Intensive Software Technologies and Applications (DISTA)
Linnaeus University Center
Linnaeus University Centre for Data Intensive Sciences and Applications (DISA)
More about the project
Artificial Intelligence is becoming an integral component in many modern software systems, necessitating the development of robust software architectures fitting machine learning components. The development of Machine Learning Operations (MLOps) architectures and standards is pivotal in ensuring smooth collaboration between software development practices and AI-driven approaches, facilitating efficient integration, deployment and inference of these AI models.
This Ph.D. project aims to define, implement, and evaluate MLOps architectures within different industrial settings. In these diverse contexts, factors such as model tracking, automatic deployment, monitoring of models, and generalisable design over architectures play a significant role in achieving seamless AI integration. Various industrial settings often present unique challenges and constraints, including varying hardware capabilities and demanding performance requirements.
In response to these challenges, the project will also investigate the optimization and acceleration of machine learning models and their respective software integrations. This multifaceted approach ensures that the MLOps architectures developed are not only adaptable but also capable of delivering high-performance and fast inference AI solutions in an industrial environment.
Throughout the research journey, an iterative process will be employed to refine and enhance the state of the art in MLOps and performance optimization. By doing so, this project aims to contribute to the development of adaptable, efficient, and scalable MLOps solutions that empower the seamless integration of AI models across diverse industrial software ecosystems.
The project is part of the research in the Data Intensive Software Technologies and Applications (DISTA) and Linnaeus University Centre for Data Intensive Sciences and Applications (DISA) research groups.
