Towards Green DevOps: Measuring and Improving Sustainability in AI-Enabled Microservices Architectures
Journal of Sustainable Software Engineering and Information Systems · 16.11.2026 · 2 (2) · 125-143
DOI: 10.58723/jsseis.v2i2.260 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Sustainable Software Engineering and Information Systems, online erschienen am 16.11.2026. Verfasst von Likinyo Ayorinde Paul und 4 weiteren Autor:innen.
Abstract (Originalsprache)
Background
The growing deployment of artificial intelligence models through cloud-native microservices has increased the energy consumption, resource demand, and operational cost of production software. However, sustainability is rarely incorporated into continuous integration and continuous delivery pipelines as a measurable software-quality attribute.
Aims
This study aims to design and evaluate a Green DevOps framework that enables development and operations teams to measure, compare, and improve the sustainability of AI-enabled microservices.
Methods
The framework integrates energy consumption, estimated operational carbon emissions, CPU and memory utilization, storage activity, cloud expenditure, latency, reliability, and predictive performance into CI/CD and runtime-observability processes. A controlled prototype evaluation examined four deployment configurations using a containerized document-classification service.
Result
Compared with the static baseline, the optimized configuration reduced estimated energy consumption and operational carbon intensity by 57.1%, CPU consumption by 50.0%, and cloud cost per 1,000 successful requests by 51.2%. It improved p95 latency from 145 to 115 ms and throughput from 100 to 109 requests/s; macro-F1 changed from 0.918 to 0.914.
Conclusion
Coordinated optimization of scaling, orchestration, caching, and model precision can improve environmental and economic efficiency without materially compromising service quality, while CI/CD sustainability gates can make this evidence actionable.
Zitieren
APA
Paul, L. A., Nurfitriani, R., John-Dewole, A. T., Jyothi, V. E., & Nagamani, T. (2026). Towards Green DevOps: Measuring and Improving Sustainability in AI-Enabled Microservices Architectures. Journal of Sustainable Software Engineering and Information Systems, 2(2), 125-143. https://doi.org/10.58723/jsseis.v2i2.260
Vancouver
Paul LA, Nurfitriani R, John-Dewole AT, Jyothi VE, Nagamani T. Towards Green DevOps: Measuring and Improving Sustainability in AI-Enabled Microservices Architectures. Journal of Sustainable Software Engineering and Information Systems. 2026;2(2):125-143. doi:10.58723/jsseis.v2i2.260
BibTeX
@article{Paul2026,
author = {Paul, Likinyo Ayorinde and Nurfitriani, Ressti and John-Dewole, A. T and Jyothi, V. Esther and Nagamani, Tenali},
title = {Towards Green DevOps: Measuring and Improving Sustainability in AI-Enabled Microservices Architectures},
journal = {Journal of Sustainable Software Engineering and Information Systems},
year = {2026},
volume = {2},
number = {2},
pages = {125--143},
doi = {10.58723/jsseis.v2i2.260}
}
EUVIMED