Research area

Human–AI Interaction

Research on how interface choices shape the way people understand and interact with AI systems.

LLM Interface Design - Node Tree

The Node-tree interface turns a linear AI conversation into an interactive hierarchy, helping people preserve context, revisit branches, and explore ideas without losing their place.

Read paper ↗

This work was done at SAP.

LLM Modality Design

Ongoing research examining how different ways of presenting AI-generated advice shape people’s interaction with and evaluation of that advice.

HARP (Human–AI Research Platform)

HARP lets researchers run controlled studies with live, configurable AI agents while capturing how people compose, revise, and hesitate over prompts-not only what they ultimately submit.

Read paper ↗

This work was done at SAP.

Mediating Cognitive Biases

Research investigating how often cognitive biases shape workplace decisions and whether LLMs can help people recognize and avoid them.

Read paper ↗

This work was done at SAP.

The Impact of Confidence Ratings on User Trust in Large Language Models

A comparative study examined how user-facing confidence ratings, confidence factors, and prompt suggestions shape trust and task effectiveness when people work with LLMs.

Read paper ↗

This work was done at SAP.