As Large Language Models (LLMs) continue to influence software engineering (SE), the integration of software design knowledge becomes increasingly important for enhancing LLM applications. This talk explores the significant role that software design play in improving the performance of LLM-based SE techniques. I will begin by discussing prevalent challenges in software design within the industry, which sets the context for understanding the transformative impact of LLMs on SE practices. Through specific examples, I will demonstrate the critical importance of incorporating software design knowledge in tasks such as code generation, test generation, and beyond, highlighting how these integrations lead to more effective and robust software solutions. Furthermore, the presentation will explore how LLMs are reshaping research in software design, fostering new methodologies and innovative approaches. This discussion aims to emphasize that a deep understanding of software design is not merely beneficial but essential for advancing LLM applications in software engineering, ensuring that these technologies meet the complex demands of modern software development.

Xin Xia is a Chief Expert in Software Engineering Application Technologies at Huawei Technologies, China. His research interests encompass artificial intelligence (AI) and software engineering (SE), mining software repositories, and empirical software engineering. Xin has authored over 340 papers and has received 15 best or distinguished paper awards, including eight ACM SIGSOFT Distinguished Paper Awards for his contributions at ASE (2018-2021), ICPC (2018, 2020), ICSE (2024), and MSR (2024). He was awarded the ACM SIGSOFT Early Career Researcher Award in 2022 and is an ACM Distinguished Member.