SE4AS aims to consolidate a scientific and technical agenda by bringing together researchers and industry practitioners interested in Software Engineering for Agentic Systems. The workshop seeks to promote a structured exchange of experiences, challenges, and solutions, contributing to:
Identification of open research challenges
Consolidation of a common vocabulary in Agentic Engineering
Discussion of architectural and methodological approaches
Encouragement of inter-institutional collaborations
Development of a research agenda for the coming years
Topics of Interest
Software engineering techniques and methodologies for agentic systems
Requirements engineering for agentic systems
Architectures for agentic systems (e.g., single-agent, multi-agent, hierarchical)
Evaluation and benchmarking of agentic systems
Testing, validation, and verification of agentic systems
Observability, monitoring, and debugging of agentic systems
Tool use and integration in agent workflows
Memory, planning, and reasoning mechanisms
Coordination, communication, and collaboration in agentic systems
Human-in-the-loop and hybrid agentic systems
Safety, ethics, and governance of agentic systems
Empirical studies and industrial experiences with agentic systems
Applications of agentic systems in a range of areas, including data science, education, healthcare, and defense
