TY - GEN
T1 - An Analysis of MLOps Architectures: A Systematic Mapping Study
AU - Amou Najafabadi, Faezeh
AU - Bogner, Justus
AU - Gerostathopoulos, Ilias
AU - Lago, Patricia
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
PY - 2024
Y1 - 2024
N2 - Context. Despite the increasing adoption of Machine Learning Operations (MLOps), teams still encounter challenges in effectively applying this paradigm to their specific projects. While there is a large variety of available tools usable for MLOps, there is simultaneously a lack of consolidated architecture knowledge that can inform the architecture design.Objective. Our primary objective is to provide a comprehensive overview of (i) how MLOps architectures are defined across the literature and (ii) which tools are mentioned to support the implementation of each architecture component. Method. We apply the Systematic Mapping Study method and select 43 primary studies via automatic, manual, and snowballing-based search and selection procedures. Subsequently, we use card sorting to synthesize the results. Results. We contribute (i) a categorization of 35 MLOps architecture components, (ii) a description of several MLOps architecture variants, and (iii) a systematic map between the identified components and the existing MLOps tools. Conclusion. This study provides an overview of the state of the art in MLOps from an architectural perspective. Researchers and practitioners can use our findings to inform the architecture design of their MLOps systems.
AB - Context. Despite the increasing adoption of Machine Learning Operations (MLOps), teams still encounter challenges in effectively applying this paradigm to their specific projects. While there is a large variety of available tools usable for MLOps, there is simultaneously a lack of consolidated architecture knowledge that can inform the architecture design.Objective. Our primary objective is to provide a comprehensive overview of (i) how MLOps architectures are defined across the literature and (ii) which tools are mentioned to support the implementation of each architecture component. Method. We apply the Systematic Mapping Study method and select 43 primary studies via automatic, manual, and snowballing-based search and selection procedures. Subsequently, we use card sorting to synthesize the results. Results. We contribute (i) a categorization of 35 MLOps architecture components, (ii) a description of several MLOps architecture variants, and (iii) a systematic map between the identified components and the existing MLOps tools. Conclusion. This study provides an overview of the state of the art in MLOps from an architectural perspective. Researchers and practitioners can use our findings to inform the architecture design of their MLOps systems.
KW - Architecture
KW - Components
KW - Machine Learning Operations
KW - MLOps
KW - Systematic Mapping Study
KW - Tools
UR - https://www.scopus.com/pages/publications/85203582483
UR - https://www.scopus.com/inward/citedby.url?scp=85203582483&partnerID=8YFLogxK
UR - https://doi.org//10.5281/zenodo.11067770
U2 - 10.1007/978-3-031-70797-1_5
DO - 10.1007/978-3-031-70797-1_5
M3 - Conference contribution
AN - SCOPUS:85203582483
SN - 9783031707964
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 69
EP - 85
BT - Software Architecture
A2 - Galster, Matthias
A2 - Scandurra, Patrizia
A2 - Mikkonen, Tommi
A2 - Oliveira Antonino, Pablo
A2 - Nakagawa, Elisa Yumi
A2 - Navarro, Elena
PB - Springer Science and Business Media Deutschland GmbH
T2 - 18th European Conference on Software Architecture, ECSA 2024
Y2 - 3 September 2024 through 6 September 2024
ER -