About Me

Hi! I'm Reza Shahriari, a PhD candidate in Computer Science at the University of Florida, advised by Dr. Eric Ragan.

My research addresses a central challenge in human–AI interaction: while Explainable AI (XAI) methods aim to improve human understanding by exposing model behavior, the resulting information can itself become overwhelming to interpret, particularly in multimodal systems that combine multiple forms of information. At the same time, AI outputs are often imperfect and require human inspection for validation and correction.

I design multimodal interaction techniques that support how users understand, inspect, and improve AI systems. Drawing on methods from human–computer interaction (HCI) and data visualization, I investigate interactive systems that help users parse complex AI outputs, reason about system behavior, identify inconsistencies, and provide human-in-the-loop (HITL) feedback within human-centered AI. To study these questions, I conduct controlled experiments using both quantitative and qualitative methods to examine human performance, interaction behavior, and the effectiveness of interaction techniques for improving AI systems.