A Review on Architectures and Techniques of Generative AI Frameworks for Intelligent Cloud

Abstract

Cloud computing has become one of the most important tools for making computing services scalable, flexible, and affordable. But, there are significant obstacles in the areas of operational automation, security, monitoring, and fault diagnosis due to the increasing complexity of cloud infrastructures, dynamic workloads, multi-cloud settings, and remote applications. Automated decision support, natural language interaction, and knowledge-driven automation are all made possible by new possibilities made possible by advances in Generative Artificial Intelligence (GenAI) and Large Language Models (LLMs), which have created new paths for intelligent cloud operations. Here, takes a close look at the Generative AI frameworks' intelligent cloud operations and analyzes their structures and approaches in detail. Afterward, the fundamentals of cloud operations are covered, along with the current obstacles, Generative AI, core foundation models, and LLMs. Following that, it delves into a number of noteworthy generative architectures, including Transformers, GANs, VAEs, and Diffusion Models, along with well-known frameworks such as LangChain, LlamaIndex, AutoGen, CrewAI, and Semantic Kernel. Agentic AI, Retrieval-Augmented Generation (RAG), and Prompt Engineering are among the enabling techniques that are covered. Finally, the literature review highlights the present research trends, gaps, and potential future directions for autonomous, secure, and intelligent cloud administration.