International Workshop on Agentic AI for Next-Generation Software Development (AgenticDev 2026)

Co-located with ASE 2026 Automated Software Engineering Conference

Recent advances in Large Language Models (LLMs) and AI systems are enabling a new paradigm in Software-AI Engineering: Agentic AI, where multiple intelligent agents collaborate with developers and with each other across the software lifecycle. Moving beyond isolated AI tools. This paradigm envisions coordinated ecosystems of AI agents capable of reasoning, planning, and acting in complex development environments.

The AgenticSE 2026 workshop aims to explore this emerging vision, inspired by major research initiatives that investigate multi-agent collaboration, human-AI interaction, and integrated AI-supported software engineering platforms. The workshop will provide a forum to discuss how agentic systems can enhance productivity, improve software quality, and enable more adaptive and trustworthy development processes.



Workshop Goals

The workshop aims to:

Foster discussion on the foundations and challenges of Agentic AI in Software Engineering
Bridge research and practice across academia and industry
Build a community around next-generation, AI-native software development


Topics of Interest

We invite original research papers, vision papers, and real-world application studies addressing related aspects including, but not limited to:

Architectures for Agentic AI systems in software engineering
Multi-agent collaboration, coordination, and orchestration
Human-AI collaboration and human-in-the-loop development
Agent-based support for requirements engineering, coding, testing, and maintenance
LLM-based agents and prompt/program synthesis techniques
Trustworthiness, verification, and validation of AI agents
Explainability, transparency, and accountability in Agentic Systems
Integration of Agentic AI into existing development workflows and tool-chains
Applications in domains such as Robotics, Cyber-Physical Systems (CPS), Edge-AI and industrial software
Benchmarking, evaluation methods, and empirical studies of Agentic Systems
Open-source frameworks and platforms for Agent-based Software Engineering


Submission Types

We welcome the following types of submissions:

Full papers (up to 10 pages): mature research contributions
Short papers (up to 5 pages): work-in-progress, vision, or position papers
Demo/Tool papers (up to 5 pages): practical systems and prototypes evaluation
All submissions must follow the ASE formatting guidelines and will be peer-reviewed based on relevance, originality, technical quality, and clarity. In particular, papers should:

be written in English;

be in PDF format and conform to the ACM Proceedings Template. LaTeX users must use the \documentclass[sigconf,review,anonymous]{acmart} option;

not exceed 5 pages (short papers) or 10 pages (regular papers). Up to two additional pages containing ONLY references are permitted.