Todorov Iliev, Dinev Pavlov, 2026. "Agentic Cloud Intelligence: Architectures for Autonomous Reasoning, Planning, and Self-Optimization in Distributed Cloud Ecosystems", International Journal of Cloud Systems and Applications (IJCSA) 1(1): 53-69.

Agentic Cloud Intelligence is the next-generation paradigm in which autonomous AI agents employ distributed cloud computing infrastructures to build intelligent adaptive self-managing cloud ecosystems. Today, traditional cloud management approaches continue to struggle under the onslaught of dynamic workloads, resource complexity and real-time operational demands due to increasing dependence on cloud-native applications and big data analytics, IoT devices and edge computing services in modern organizations. Traditional cloud systems are highly dependent on established automation principles and human input, constraining their ability to react in real time to the variables of an evolving atmosphere. These limitations of contemporary cloud computing are recognized by Agentic Cloud Intelligence, which lets agents autonomously reason, plan, learn and optimize the operation with little to no human input.
Agentic Cloud Intelligence architecture combines state-of-the-art technologies like machine learning, deep learning, reinforcement learning, large language models, knowledge graphs and multi-agent systems. These technologies provide automated ways for intelligent agents to sense the environmental context, reason through contextual information, make decisions, and autonomously perform actions. Cloud agents have the ability to perform intelligent functions through cognitive capabilities such as reasoning, planning, memory management and adaptive learning to continuously monitor infrastructure performance, make predictions of resource requirements, detect anomalies and optimize service delivery. Memory-augmented architectures additionally enable retention of knowledge and contextual understanding, enabling the agent to learn from decisions progressively.

1. No References