Presentation at ECSA 2024: A Systematic Mapping Study of MLOps Architectures

This work was co-authored by Justus Bogner, Ilias Gerostathopoulos, Patricia Lago, and Faezeh Amou Najafabadi. It offers a deep dive into how MLOps architectures are described in the academic literature, presenting 35 essential architectural components and several variants. The partners have also provided a comprehensive mapping of the tools and technologies associated with these components, highlighting the evolving landscape of machine learning operations (MLOps). The conference proved to be an excellent platform for engaging with both researchers and practitioners. Partners exchanged valuable insights on the challenges and advancements in MLOps, further enriching the ongoing discussions around this critical area in the software architecture domain. A highlight of our participation was receiving honorable mention in the Research Papers track, a testament to the relevance and impact of our work. For those interested in delving deeper, you can access the full paper here.