AI-Native Enterprise Architectures: Design Principles and Patterns for Next-Generation Intelligent Systems
TL;DR
This research proposes an AI-native enterprise architecture framework that identifies essential design principles and reusable architectural patterns for next-generation intelligent systems and evaluates AI-native design patterns, including intelligent agents, Retrieval-Augmented Generation (RAG), AI orchestration, and human-AI collaboration patterns.
Abstract
The emergence of intelligent computing paradigms has accelerated the transformation of conventional enterprise systems into AI-native architectures capable of autonomous reasoning, continuous learning, and adaptive decision-making. However, existing enterprise architectures provide limited guidance for integrating AI capabilities, autonomous agents, and governance mechanisms into a unified framework. This research proposes an AI-native enterprise architecture framework that identifies essential design principles and reusable architectural patterns for next-generation intelligent systems. A secondary analysis methodology is applied by examining existing academic literature, industry frameworks, and AI-enabled enterprise architectures. The findings identify six major architectural layers, including computing, storage, data, AI intelligence, application, and Model-as-aService (MaaS) governance layers. The research further evaluates AI-native design patterns, including intelligent agents, Retrieval-Augmented Generation (RAG), AI orchestration, and human-AI collaboration patterns. The proposed framework supports scalable, adaptive, secure, and trustworthy enterprise systems by embedding intelligence throughout the architecture lifecycle.