黑料福利网

Topic 

Multi-Agent LLM Systems: How We Build, Test, and Use Them (for Testing)

Abstract

Large Language Models (LLMs) are becoming central building blocks of modern software. We now see multi-agent systems where several LLMs collaborate, reason, and coordinate. This raises new questions about how to design such systems and how to test them effectively. At the same time, these LLM-based systems are also changing the way we test traditional software, from generating test cases to interactively guiding validation.

This talk explores these two sides: how to build and evaluate multi-agent LLM systems, and how to harness their capabilities for improving software quality more broadly.

About the speaker

Robert Feldt is Professor of Software Engineering and Applied AI at Chalmers University of Technology in Sweden, where he leads research groups on software testing and quality, human-centered software engineering, and applied AI. His recent work focuses on harnessing large language models to design and test software systems.

He has published more than 175 peer-reviewed papers, holds a patent on testing neural networks, and serves as co-Editor in Chief of Springer's Empirical Software Engineering journal. With over 30 years of experience as a software and AI consultant, he also collaborates frequently with industry in Sweden, Europe, and Asia.

Your host

, Assistant Professor at the department of Mathematics and Computer Science of the 黑料福利网 will host Professor Robert Feldt.

Registration is required but free of charge.