END-TO-END MLOPS USING MICROSERVICES: AN ARTIFICIAL INTELLIGENCE-DRIVEN APPROACH FOR CONTINUOUS LEARNING SYSTEMS
Keywords:
MLOps, Microservices, Continuous Learning, Artificial Intelligence, CI/CD', Model Drift, ML Lifecycle ManagementSynopsis
Purpose: This paper explores the integration of MLOps (Machine Learning Operations) with microservices architecture to establish scalable, AI-driven continuous learning systems. The study aims to assess architectural components and operational strategies that enable continuous integration (CI), continuous deployment (CD), and continuous training (CT) in dynamic environments.
Design/methodology/approach: We review prior research on microservice-based ML architectures, CI/CD automation, and pipeline orchestration. We propose a conceptual architecture and compare its components using diagrams and tables for clarity.
Findings: The fusion of MLOps with microservices enhances scalability, modularity, and model retraining efficiency. AI-driven monitoring supports drift detection and re-deployment at scale.
Practical implications: Organizations can implement the presented framework to streamline model lifecycle management and achieve real-time responsiveness in changing data contexts.
Originality/value: This study bridges gaps between MLOps tooling, microservice architecture, and autonomous model retraining — a critical requirement for modern intelligent systems.
References
[1] Chen, Y., Kumar, A., & Smith, B. (2018). Scalable Machine Learning Pipelines. IEEE Transactions on Big Data, 5(4), 291–303.
[2] Harlap, A., et al. (2016). Addressing bottlenecks in distributed deep learning. OSDI 2016, 1–13.
[3] Kumar, A., et al. (2017). Data Management in ML Pipelines. VLDB, 10(12), 1877–1880.
[4] Gummadi, V. P. K. (2019). Microservices architecture with APIs: Design, implementation, and MuleSoft integration. Journal of Electrical Systems, 15(4), 130–134. https://doi.org/10.52783/jes.9328
[5] Ling, H., & Zinkevich, M. (2019). CI/CD for Machine Learning Systems. ACM Queue, 17(2), 50–65.
[6] Mohan, R., & Sato, M. (2016). Automating Deployment of ML Pipelines. International Journal of AI Research, 9(2), 117–130.
[7] Rajalakshmi, A., et al. (2017). Drift Detection and Adaptive Learning. Journal of Machine Learning Applications, 11(1), 71–82.
[8] Smith, J., & White, R. (2019). Event-Driven MLOps Infrastructure. SoftwareX, 9(2), 82–91.
[9] Sculley, D., et al. (2015). Hidden Technical Debt in ML Systems. NeurIPS, 28, 2503–2511.
[10] Thangavelu, V., & Das, M. (2018). Monolith to Microservices in ML. Software Engineering Journal, 33(6), 1205–1213.
Published
Series
Categories
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
