AI-Driven Orchestration Framework for Cloud Computing Platforms
Keywords:
AI orchestration, cloud computing, deep reinforcement learning, LSTM workload prediction, multi-cloud management, federated learning, Kubernetes, resource schedulingAbstract
The rise in the number of workloads in the cloud computing has been exponentially increasing and the growing complexity of multi-cloud implementations require the use of intelligent orchestration mechanisms that can no longer be achieved through standard rule-based automation. The AI-driven Orchestration Framework (ADOF) offered in this paper is a cloud computing platform that combines Long Short-Term Memory (LSTM) networks to predict the workload, Deep Q-Network (DQN) reinforcement learning to schedule the resources dynamically, and a federated privacy engine to train the cross-tenant models in a secure way. The framework proposed will consider a five-layer hierarchical structure which will include resource pooling, infrastructure abstraction, orchestration control, AI intelligence, and application interface layers. An experimental assessment of a multi-cloud testbed of heterogeneous clusters (which comprises heterogeneous resources and operational frameworks) proves that ADOF can lead to a 35.9 percent increase in the utilization of the resources, a 43.1 percent decrease in the average response time, and a 26.6 percent drop in the cost of operations as compared to traditional rule based orchestration baselines. The framework also demonstrates strong SLA compliance in the conditions of bursty workload and seamless integration with the existing frontend-back backend AI systems. These findings highlight the potential of the transformative nature of integrating learning-based intelligence into the cloud orchestration control planes.
References
1. Bitsakos, C., Tsoumakos, D., Konstantinou, I., & Koziris, N. (2025). DInos: A deep reinforcement learning approach to generalizable autoscaling in stateless cloud applications. In R. Wrembel, G. Kotsis, A. M. Tjoa, & I. Khalil (Eds.), Database and Expert Systems Applications. DEXA 2025. Lecture Notes in Computer Science, vol 16046. Springer. https://doi.org/10.1007/978-3-032-02049-9_20
2. Clarifai. (2025, January 13). Hybrid cloud orchestration explained: AI-driven efficiency, cost control. https://www.clarifai.com/blog/hybrid-cloud-orchestration
3. CloudServ AI. (2025). AI-driven cloud orchestration: Revolutionizing cloud management in 2025. https://cloudserv.ai/ai-driven-cloud-orchestration-revolutionizing-cloud-management-in-2025/
4. Frontiers in Computer Science. (2025). Machine learning-based cloud resource allocation algorithms: A comprehensive comparative review. Frontiers in Computer Science, 7, 1678976. https://doi.org/10.3389/fcomp.2025.1678976
5. Mahimalur, R. K. (2025). Machine learning approaches for resource allocation in heterogeneous cloud-edge computing. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 11(2), 2739–2748. https://doi.org/10.32628/CSEIT25112758
6. Mishra, P., Hans, S., Saha, D., & Moogi, P. (2024). Optimizing cloud workloads: Autoscaling with reinforcement learning. In 2024 IEEE 17th International Conference on Cloud Computing (CLOUD) (pp. 217–222). IEEE. https://doi.org/10.1109/CLOUD62652.2024.00035
7. Mohammadi, S., Balador, A., Sinaei, S., & Flammini, F. (2024). Balancing privacy and performance in federated learning: A systematic literature review on methods and metrics. Journal of Parallel and Distributed Computing, 192, 104918. https://doi.org/10.1016/j.jpdc.2024.104918
8. Parra-Ullauri, J. M., Zhou, X., Moazzeni, S., Hussain, R., Vasilakos, X., & Simeonidou, D. (2024). kubeFlower: A privacy-preserving framework for Kubernetes-based federated learning in cloud–edge environments. Future Generation Computer Systems, 157, 1–17. https://doi.org/10.1016/j.future.2024.03.041
9. Pintye, I., Kovács, J., & Lovas, R. (2024). Enhancing machine learning-based autoscaling for cloud resource orchestration. Journal of Grid Computing, 22, 68. https://doi.org/10.1007/s10723-024-09783-1
10. Rahmani, S. (2025). Cloud-native predictive scaling framework using ML-driven workload forecasting for enterprise IT systems. Journal of Multidisciplinary Knowledge, 5(2), 216–225. https://doi.org/10.36676/jmk.v5.i2.100
11. Sensors — MDPI. (2023). Deep reinforcement learning for workload prediction in federated cloud environments. Sensors, 23(15), 6911. https://doi.org/10.3390/s23156911
12. Springer Nature. (2024). Deep reinforcement learning-based methods for resource scheduling in cloud computing: A review and future directions. Artificial Intelligence Review, 57, 162. https://doi.org/10.1007/s10462-024-10756-9
13. Talluri, M., & Rachamala, N. R. (2023). Orchestrating frontend and backend integration in AI enhanced BI systems. International Journal of Intelligent Systems and Applications in Engineering, 11(1), 679–688. ISSN: 2147-6799. www.ijisae.org
14. Wang, H., Wang, Q., & Ding, Y. (2024). Privacy-preserving federated learning based on partial low-quality data. Journal of Cloud Computing, 13, 62. https://doi.org/10.1186/s13677-024-00618-8
15. Wang, Y., & Li, J. (2025). Intelligent resource allocation optimization for cloud computing via machine learning. arXiv preprint arXiv:2504.03682. https://arxiv.org/abs/2504.03682
