Article on AI Agents and Software Architectures published in IEEE Software
- Post by: admin
- March 30, 2026
- No Comment
Making Sense of AI Agents Hype: Adoption, Architectures, and Takeaways from Practitioners
Ruoyu Su, Matteo Esposito, Roberta Capuano, Rafiullah Omar, June Sallou, Henry Muccini, Davide Taibi
FrAmeLab researchers Roberta, Rafiullah and Henry, together with collaborators from the University of Oulu, Wageningen University, and the University of Southern Denmark, published the article “Making Sense of AI Agents Hype: Adoption, Architectures, and Takeaways from Practitioners” in IEEE Software. The work presents a large-scale empirical study on the industrial adoption of AI agent-based systems and investigates how organizations are integrating LLM-based AI agents into real-world software architectures. The study analyzes 138 practitioner talks selected from a dataset of 234 international conference presentations and technical sessions focused on AI agents. Through qualitative analysis supported by LLMs and human validation, the research identifies recurring architectural strategies, coordination patterns, implementation technologies, and adoption challenges emerging in industrial practice.
The paper highlights how companies are transitioning from experimentation to production-ready agentic systems, emphasizing key aspects such as orchestration, memory management, workflow coordination, scalability, observability, and governance. The results also show that successful agentic systems rely on explicit architectural structures, including task decomposition, modular coordination, shared context, and domain-specific interaction patterns.
The article provides practical guidance for researchers and practitioners working on AI-driven software systems and contributes a consolidated overview of industrial trends in agentic architectures, frameworks, and deployment strategies.
https://ieeexplore.ieee.org/document/11474799