Talk Title: Lessons from the Trenches in Building Agents for Software Development

Talk Abstract: Over the past year, AI agents have rapidly developed from a curiosity to a core part of many development workflows. While this development may seem like it was inevitable, it was actually built on a series of rapid technological advances, many built with the assistance of software development agents themselves. In this talk I will talk about several key technologies enabling software-based agents, including the simple but powerful tools developed to provide models with the interfaces that they need, rigorous evaluation benchmarks, and training of agentic models. Further, I will talk about some developing research topics, such as encouraging human-agent interaction, agent memory, and task decomposition.

Bio: Graham Neubig is an associate professor at the Language Technologies Institute of Carnegie Mellon University and Chief Scientist at OpenHands. His research focuses on large language models, including both fundamental advances in model capabilities and applications to tasks such as software development. His final goal is that every person in the world should be able to communicate with each-other, and with computers in their own language. He also contributes to making NLP research more accessible through open publishing of research papers, advanced NLP course materials and video lectures, and open-source software, all of which are available on his web site.