Introduction

• Michel Jouvin (Université Paris-Saclay (FR))
• Nicole Skidmore (University of Warwick)
• Claire Antel (Universite de Geneve (CH))

Room: 160/R-009

Balancing accuracy, performance, and maintainability in research software Room: 160/R-009 Abstract: Technical debt refers to short-term optimizations that may lead to longer-term consequences in the software project. Spoiler alert - it happens in research software as well. In this talk, I will share some findings from studies my group has done into the nature of technical debt in research software projects, both large and small. Key is to do tradeoff analysis of your important project qualities, like accuracy or performance. Using a combination of interviews, surveys, and quantitative analysis, we present some insights into how different projects perceive technical debt, how they resolve it, and how technical debt has a different flavour in research software. I will conclude with some ongoing work on using AI to fill in domain knowledge gaps. Speaker: Neil Ernst, PhD, is an associate professor at the University of Victoria in the Department of Computer Science and Director, Matrix Institute for Data Science. He is a world-leading researcher in software architecture and requirements. His research focuses on building next generation software systems. He leverages past experience consulting with large government stakeholders and empirical datasets on software development and analysis. Current projects include technical debt in scientific software, climate informatics, and the impacts of AI on human aspects of software development.