In recent years, Large Language Models (LLMs) have emerged as a transformative technology, opening new opportunities across various fields, including Information Systems design. While LLMs excel in Natural Language Processing tasks, such as text translation and summarization, their potential in software architecture design and, in particular, in service-oriented solutions, is still underexplored.

This workshop aims to provide a forum for innovative proposals striving to integrate LLMs in the landscape of Service-Oriented Architectures and Systems. Of particular interest are the benefits delivered by the adoption of LLMs in improving the design of service-oriented architectural solutions (and their impact on efficiency, accuracy, and scalability) as well as their use in tasks such as service discovery and composition.

Topics of interest (main, but not limited to):
• Usage of LLMs to improve efficiency, accuracy, and scalability of software architectures, with a particular concern on service-oriented solutions
• Approaches for service discovery and composition with LLMs
• Use of LLMs for the improvement of Service-Oriented Architectures and Systems design
• Interoperability and integration of services in Service-Oriented Architectures and Systems with LLMs
• Practical implementations and evaluation of LLMs for service interoperability
• Novel approaches for real-world applications of LLMs in Service-Oriented Architectures and Systems

Workshop Organizers
• Massimiliano Garda (University of Brescia, Italy)
• Ada Bagozi (University of Brescia, Italy)
• Ilche Georgievski (University of Stuttgart, Germany)

Advisory Board
• Massimo Mecella (University of Rome "La Sapienza", Italy)
• Monique Snoeck (KU Leuven, Belgium)
• Barbara Pernici (Politecnico di Milano, Italy)
• Devis Bianchini (University of Brescia, Italy)

Submission instructions
The workshop accepts regular papers (up to 12 pages) and short/position papers (from 6 up to 8 pages). Ongoing research work is also welcome. Page limit for regular and short/position papers includes all text, figures, references, and appendices.

Papers should be submitted as PDF, following Springer's LNCS/LNBIP format guidelines, using EasyChair (https://easychair.org/my/conference?conf=caise2026) and choosing the track "Workshops - LLM-SOA".

For any inquiries related to this workshop, please feel free to contact us at: llm-soa2026@easychair.org