Overview

About the workshop

LLM agents increasingly collaborate and utilize external tools to address complex user requests. This emerging paradigm demands vast volumes of data for training, testing, reasoning, memory, and execution to ensure accurate and timely responses and actions. Despite the clear benefits, there remains a pressing need for a more rigorous characterization of software engineering constructs to effectively support LLM agents, particularly in areas such as orchestration, data management, knowledge augmentation, and planning. This workshop explores novel solutions to these challenges from the joint perspectives of software engineering and big data.

Aims

The goal of this workshop is to bring together original, high-quality contributions on software engineering challenges and applied solutions for agentic AI and its synergy with big data, spanning architecture, development, verification, observability, agent-tool protocols, and resource management.

Topics
Topics of interest include, but are not limited to:
  • Agentic big data frameworks and systems.
  • Requirements engineering for agentic systems.
  • Design and architecture of LLM-based multi-agent systems.
  • Development processes and data management in agentic systems.
  • Code and other artifact generation assisted by software agents.
  • Standardized agent-tool protocols (e.g., MCP, A2A).
  • Human-in-the-loop in agentic systems.
  • Conversational agents and chatbot frameworks.
  • Structured and unstructured data manipulation for agentic recommender systems.
  • Reproducibility and traceability of agentic systems development.
  • Testing, verification, validation, and LLM-based multi-agent consensus mechanisms.
  • Evaluation approaches and frameworks including LLMs as judges and LLM-based qualitative and quantitative metrics.
  • Observability, accountability, audit trails, and reproducibility in LLM-based multi-agent systems.
  • Empirical studies of multi-agent frameworks (e.g. LangChain, AutoGen, CrewAI).
  • Domain-specific applications, including data science, spatial-temporal, and agentic recommender systems.
  • Self-adaptation and other self-* properties in the context of agentic AI.
  • Real-time or dynamic agentic execution, visualization, and workflows.

Submissions

Call For Papers
We invite original research papers on Software Engineering for Agentic AI.
  • Short papers: 3-4 pages, including references
  • Regular papers: 8-10 pages, including references
Papers should follow the IEEE Conference 2-column format.

Important dates
  • Submission deadline: October 10, 2026.
  • Notification of paper acceptance to authors: October 31, 2026 November 5, 2026
  • Camera-ready of accepted papers: November 14, 2026 November 26, 2026
  • Workshop / conference dates: December 14 - 17, 2026
Schedule
Time Title Authors
TBA Engineering Reliable Commit Gates for Agentic AI: Cost-Aware Verification Portfolios under Common-Mode Data Failures Zihao Zheng, Baichuan Li, Junyi Yao, and Jiayu Long
TBA Who Grades the Grader? The Verification Gap in Self-Testing Code Agents Brian Sam-Bodden, Jyoti Vasudev, Josh Meyers, and Eric Kasper
TBA Beyond the Black Box: Concurrency-Correct Latency Analytics over Large-Scale LLM-Agent Execution Telemetry Yash Parikh and Nikhil Patne
TBA What Do Test Assertions Rest On? Measuring Cross-Stage Provenance in Agentic BDD Pipelines Chieh-Hsuan Lu
TBA Generation Time Security Safeguards in Agent Instruction Files Chandra Inguva and Nabita Penmetsa
TBA ClaimScope: A Repository-Level Reanalysis of Artifact-Size Reporting in Agentic Pull Requests Moon Hwan Lee
TBA Testing the Grader: Canaries and Fault Injection for Agent-Written Lean Code Robert Sneiderman
TBA Can Agents Trust a Replication Verifier? Epistemic Contracts for Scientific Agent Tools Luxin Zhang
TBA Regression Testing Model Upgrades in Policy-Bound Enterprise Agents Ravikanth Somireddy, Anshul Singhal, Praveen Chinnappa, and Anshuman Jain
TBA Node-as-Agent: Graph Agentic Network Minghao Guo, Xi Zhu, Qingyue Jiao, Xiujin Liu, Haochen Xue, Chong Zhang, Shuhang Lin, Jingyuan Huang, Ziyi Ye, and Yongfeng Zhang
TBA SAGE-SQL: Schema-Adaptive Agentic Text-to-SQL Generation and Secure Remote SQL Execution Architecture Rajvi Desai, Raghu Katikeri, Sathishkumar Venkatachalam, Mayank Kumar Jha, Anush Jain, and Amit Vaid
TBA SPEC-CHECK: Verifying LLM Agents on Cloud Infrastructure Estimation Tasks Vijaya Peketi
TBA Separating Representation and Evidence Policy in Tool-Agent Evaluation Chengxu Bian

Contact

Organizers
  • Tales Paiva: tmellopaiva@uwaterloo.ca
  • Ivens: iportugal@uwaterloo.ca
  • Paulo Alencar: palencar@uwaterloo.ca
  • Donald Cowan: dcowan@uwaterloo.ca
PC Members
  • Alessandra Parziale, Gran Sasso Science Institute, Italy
  • Abhishek Pandey, META, USA
  • Aninditha Ramesh, DoorDash, USA
  • Arpit Thool, Virginia Tech, USA
  • Chandra Inguva, UC Berkeley, USA
  • Dinesh Besiahgari, University of Cincinnati, USA
  • Dhruv Patel, Independent Researcher, USA
  • Enrico Vicario, University of Florence, Italy
  • George Tambouratzis, Athena Research Center, Greece
  • Girish Dasarathy, Microsoft, USA
  • Giuliano Lorenzoni, University of Waterloo, Canada
  • Hadi Hasan, American University of Beirut, Lebanon
  • Ivan Compagnucci, Gran Sasso Science Institute, Italy
  • Mahdi Ghadamyari, Google, USA
  • Marc Maurer, Stealth AI Startup, USA
  • Marco Torchiano, Politecnico di Torino, Italy
  • Martin Mirchev, Eindhoven University of Technology, Netherlands
  • Navaprasad Maram, Intuit, USA
  • Nishant Ravi Shankar, Oracle, USA
  • Praveen Myakala, JP Morgan & Chase, USA
  • Rafael Kalandadze, Wandero AI, USA
  • Sarita Kumari, Amazon, USA
  • Satishkumar Masilamani, Intuiti, USA
  • Stefano Lambiase, Aalborg University, Denmark
  • Sumeet Ram, PayPal, USA
  • Vikrant Bachhav, Condor Software, USA
  • Yujun Zhao, META, USA
  • Yunqing Li, Lenovo US, USA

Previous Editions

2025