Dylan Richard

I work on applied artificial intelligence, with a particular interest in how large language models retrieve, compare and recommend information.

My current focus is on generative search, AI-assisted research and agentic workflows, especially in professional-service environments.

My background is in legal and professional services. I am interested in practical AI systems, real-world model behavior and the ways generative interfaces are changing how people discover information and make decisions.

This page documents notes, experiments and ongoing work developed alongside the Oxford Artificial Intelligence Programme at Saïd Business School, University of Oxford.

I am particularly interested in generative search: how large language models retrieve information, identify relevant entities, compare alternatives and ultimately decide what — or whom — to recommend. I am also exploring the instability of model outputs, the relationship between traditional search visibility and AI-generated recommendations, and practical ways of evaluating these systems in real-world environments.