CFP: International Workshop on Architecting Agentic AI Systems (AgentArch2027)
Feng XIA <[email protected]>
| Newsgroups | gmane.comp.hci.acm-sigchi.announce |
|---|---|
| Message-ID | <CAO_+Sr0LBuSMiTRjAhZpOY5yMFeOCnjZx0VM4MQENvi2o07tWg@mail.gmail.com> |
AgentArch 2027: International Workshop on Architecting Agentic AI Systems
Co-located with ICSA 2027 (The 24th IEEE International Conference on
Software Architecture)
Sydney, Australia | 8-12 March 2027
Website: https://agentarch.org
Submission System: https://easychair.org/conferences/?conf=icsa2027
CALL FOR PAPERS
The International Workshop on Architecting Agentic AI Systems (AgentArch
2027) invites original research contributions on the architectural
foundations of AI agents, multi-agent systems, and human-agent
collaboration. Recent advances in large language models (LLMs) have shifted
artificial intelligence from standalone models toward agentic systems
capable of reasoning, planning, tool use, memory management, and autonomous
decision making. These systems increasingly operate as complex software
systems involving multiple interacting agents and humans. However, despite
rapid progress in AI capabilities, the software architectural foundations
required to design, analyse, evaluate, and govern agentic systems remain
largely unexplored. AgentArch aims to establish software architecture as a
fundamental perspective for understanding and engineering next-generation
agentic AI systems. The workshop brings together researchers and
practitioners from software architecture, artificial intelligence, software
engineering, human-computer interaction, distributed systems, and related
disciplines to discuss emerging challenges, architectural principles,
reusable design patterns, and future research directions.
Topics of Interest
The workshop welcomes, but is not limited to, the following topics.
1) Architectures for AI Agents
- Agent architectures for reasoning, planning, and tool use
- Memory, state management, and context handling
- Prompting pipelines and modular reasoning workflows
- Integration of external tools, APIs, and knowledge sources
- Observability, monitoring, and debugging of agent behaviours
2) Multi-Agent Systems and Coordination
- Communication protocols and coordination mechanisms
- Distributed and decentralised agent architectures
- Task decomposition and collaborative problem solving
- Emergent behaviours and system-level dynamics
- Reliability, robustness, and scalability
3) Human-Agent Collaboration
- Human-in-the-loop and mixed-initiative systems
- Decision support and human-AI co-creation
- Transparency, interpretability, and traceable reasoning
- Trust, controllability, and user alignment
- Fairness, bias propagation, and responsible AI
Submission Guidelines
We solicit contributions in two categories.
- Research Papers Up to 8 pages (IEEE format)
- Short Papers Up to 4 pages (IEEE format)
All submissions must follow the IEEE Computer Society proceedings format.
Submissions must be received by the submission deadline and must not have
been published previously or be under review elsewhere while under
consideration.
Accepted papers will be published in the ICSA 2027 Companion Proceedings
and in the IEEE Xplore Digital Library.
Each submission will be reviewed by at least three members of the Program
Committee based on originality, technical quality, relevance, and potential
impact.
At least one author of each accepted paper must register and present the
paper at the workshop.
Authors may be invited to present a poster in addition to the oral
presentation.
A Best Paper Award will be presented to recognise outstanding contributions.
Important Dates (AoE)
- Paper Submission Deadline: 20 December 2026
- Notification of Acceptance: 19 January 2027
- Camera-ready Deadline: 29 January 2027
Organizers
- Ziqi Xu, RMIT University, Australia
- Ivan Compagnucci, Gran Sasso Science Institute, Italy
- Man-Fai Leung, Anglia Ruskin University, United Kingdom
- Shuo Yu, Dalian University of Technology, China
- Sajal K. Das, Missouri University of Science and Technology, USA
- Feng Xia, RMIT University, Australia
----------------------------------------------------------------------------------------
To unsubscribe from CHI-ANNOUNCEMENTS send an email to:
mailto:[email protected]
To manage your SIGCHI Mailing lists or read our polices see:
https://sigchi.org/operations/listserv/
----------------------------------------------------------------------------------------