Paper Number
ECIS2025-1965
Paper Type
SP
Abstract
Artificial intelligence (AI) and machine learning (ML) are transforming industries by enabling automation, optimization, and co-creation of business value. However, the deployment of ML models into production ecosystems and enterprise-wide information systems remains a significant challenge, partly due to the lack of standardized frameworks addressing the full ML lifecycle. While existing lifecycle models focus heavily on pre-deployment phases like data preparation and model training, critical post-deployment tasks such as monitoring, maintenance, and governance are often overlooked. To address this gap, we conduct a systematic mapping study to analyze existing ML lifecycle frameworks, identifying their limitations in post-deployment coverage. Building on these findings, we propose a research design to create a deployment-aware ML lifecycle, integrating empirical insights from multiple case studies across industries. This comprehensive framework aims to support practitioners in managing the complexities of real-world ML deployments and ensuring the sustainability of ML systems within dynamic organizational contexts.
Recommended Citation
Zall, Manuel and Thiesse, Frédéric, "Towards a Deployment-Aware Machine Learning Lifecycle" (2025). ECIS 2025 Proceedings. 4.
https://aisel.aisnet.org/ecis2025/ent_system/ent_system/4
Towards a Deployment-Aware Machine Learning Lifecycle
Artificial intelligence (AI) and machine learning (ML) are transforming industries by enabling automation, optimization, and co-creation of business value. However, the deployment of ML models into production ecosystems and enterprise-wide information systems remains a significant challenge, partly due to the lack of standardized frameworks addressing the full ML lifecycle. While existing lifecycle models focus heavily on pre-deployment phases like data preparation and model training, critical post-deployment tasks such as monitoring, maintenance, and governance are often overlooked. To address this gap, we conduct a systematic mapping study to analyze existing ML lifecycle frameworks, identifying their limitations in post-deployment coverage. Building on these findings, we propose a research design to create a deployment-aware ML lifecycle, integrating empirical insights from multiple case studies across industries. This comprehensive framework aims to support practitioners in managing the complexities of real-world ML deployments and ensuring the sustainability of ML systems within dynamic organizational contexts.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.