Multi-persona LLM System
An interactive debate platform where AI personas discuss controversial topics from diverse viewpoints. Designed and evaluated to understand how LLM-driven dialogues can reduce confirmation bias and enhance critical thinking.
Role: Lead UX Researcher
Methods: Eye-tracking, Mixed Methods, Prototype Evaluation
Transparent Fact-checking System
User study with AI-powered fact-checking system for COVID-19 related topics.
Relevance Assessment in Search Engines
User study to investigate the cognitive load when users assess relevance of the search documents.
“Good design starts with an understanding of psychology and technology. ”
— Don Norman <The Design of Everyday Things>