This paper introduces a predictive orchestration model for decentralized cloud environments. The system anticipates workload surges using historical data and proactively scales resources, ensuring service availability while reducing operational costs through smart provisioning strategies across hybrid nodes.

Felix Emmanuel Boyd, Tang Li Xun, Aanya Pranav Deshmukh, Isabelle Claire Dawson, Mateo Fernando Ruiz

This research presents a digital twin architecture integrating cloud and edge AI for infrastructure monitoring. Real-time sensor data is analyzed for anomaly detection, predictive maintenance, and operational insights, supporting resilient and intelligent urban infrastructure systems.

Samuel Noah Lambert, Chen Yu Lan, Sudeep Ranganath Pillai, Fiona Leigh Andrews, Leonardo Miguel Pacheco

This study introduces a federated learning strategy that securely fuses data across multi-tenant clouds. It ensures privacy preservation using homomorphic encryption while enabling collaborative training for AI applications with minimal latency and strong performance on diverse datasets.

Rowan Charles Mitchell, Liang Wei Zhao, Rhea Sanjana Mohan, Henry Jacob Stirling, Bianca Helena Caruso

This paper explores energy-aware VNF placement using genetic algorithms in cloud networks. By simulating biological evolution, the model optimizes resource utilization while maintaining service quality, addressing energy concerns in dynamic multi-region service function chains.

Jasper Nathaniel Ford, Yue Qin Cheng, Ritesh Anirudh Bhatia, Madeleine Iris Stevenson, Thiago Daniel Costa

This study proposes a novel framework to mitigate cold start delays in serverless deep learning applications. By using model preloading strategies and warm container rotation, it achieves low-latency inference for real-time applications in cloud-hosted inference systems.

Benjamin Marcus Irvine, Qin Zhen Rong, Arya Nikhil Varadhan, Georgia Elise Monroe, Santiago Rafael Delgado

This research introduces a multilingual cloud-based ASR system using attention-augmented transformers. It leverages self-supervised learning for robust speech understanding, enabling real-time processing and transcription of diverse languages across industries with scalable cloud deployment capabilities.

Harper Zoe Cunningham, Zhang Hui Lin, Kartik Devendran Murthy, Emily Paige Rowland, Lucas Enzo Moretti

This paper presents an ontology-driven system for automated policy inference and compliance auditing across multicloud environments. It continuously monitors configurations, detects violations, and recommends corrective actions to maintain regulatory standards in federated service infrastructures.

Sebastian Oliver Reid, Liu Mei Shan, Vaibhav Aditya Trivedi, Charlotte Jane Matthews, Andre Felipe Rojas