EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Engineering Resilient AI-Native Software Systems: A Framework for Integrating Intelligent Services, Event-Driven Architecture, and Autonomous Runtime Adaptation

Amil Uslu

Journal of Intelligent Decision Making and Information Science · 15.09.2026 · 3 (19s) · 1047-1070

DOI: 10.17762/jidmis.v3.6256 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 15.09.2026. Verfasst von Amil Uslu. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing integration of artificial intelligence into software systems is changing the principles through which complex applications are designed, deployed, and maintained. Modern AI-native applications frequently combine intelligent services with microservices, event-driven communication, cloud infrastructure, distributed data stores, and automated deployment pipelines. While this combination creates opportunities for scalability and intelligent decision-making, it also introduces new forms of uncertainty. Model behavior may vary with data and context, service dependencies may change dynamically, and runtime conditions can evolve faster than conventional release cycles can accommodate. These characteristics require software architectures that can remain dependable while continuously adapting to changing operational conditions. This article develops a resilience-centered framework for AI-native software engineering that integrates intelligent services, event-driven architecture, and autonomous runtime adaptation. The proposed perspective treats resilience as a system-level capability rather than as an isolated collection of fault-recovery mechanisms. Event-driven architecture provides a mechanism for distributing information about changes in system state, while observability establishes the evidence required to evaluate those changes. Intelligent services can then contribute to runtime interpretation and decision-making, provided that their actions remain constrained by explicit policies and architectural boundaries. Building on established research in self-adaptive software, the framework distinguishes between simple automated recovery, performance optimization, and higher-order architectural adaptation (Krupitzer et al., 2015). Particular attention is given to the relationship between autonomy and control. Autonomous behavior can improve responsiveness when systems encounter workload fluctuations, service degradation, resource constraints, or changing data conditions, but unrestricted adaptation may itself become a source of instability. The proposed framework therefore emphasizes reversible actions, evidence-based decisions, continuous feedback, policy enforcement, and graduated levels of adaptation authority. The article further examines observability, uncertainty management, security, governance, continuous delivery, and the long-term evolution of AI-native architectures. The resulting framework positions intelligent software not merely as software that contains AI capabilities, but as a continuously monitored and controlled environment in which intelligence can contribute to resilience without compromising reliability, security, or operational continuity.

Zitieren

APA

Uslu, A. (2026). Engineering Resilient AI-Native Software Systems: A Framework for Integrating Intelligent Services, Event-Driven Architecture, and Autonomous Runtime Adaptation. Journal of Intelligent Decision Making and Information Science, 3(19s), 1047-1070. https://doi.org/10.17762/jidmis.v3.6256

Vancouver

Uslu A. Engineering Resilient AI-Native Software Systems: A Framework for Integrating Intelligent Services, Event-Driven Architecture, and Autonomous Runtime Adaptation. Journal of Intelligent Decision Making and Information Science. 2026;3(19s):1047-1070. doi:10.17762/jidmis.v3.6256

BibTeX
@article{Uslu2026,
  author = {Uslu, Amil},
  title = {Engineering Resilient AI-Native Software Systems: A Framework for Integrating Intelligent Services, Event-Driven Architecture, and Autonomous Runtime Adaptation},
  journal = {Journal of Intelligent Decision Making and Information Science},
  year = {2026},
  volume = {3},
  number = {19s},
  pages = {1047--1070},
  doi = {10.17762/jidmis.v3.6256}
}
Kostenlos starten →