Human–AI Interaction
Jacaranda
How can people adapt web interfaces to fit their own tasks?

Current work
Exploring how AI can help people combine features across websites while preserving their control over a custom workspace.
We are investigating which parts of a workspace people want to shape themselves and which they want AI to handle.
Project overview reviewed Fall 2026.
Interface prototype
Jacaranda explores moving and adapting interface features across websites. Its prototypes illustrate how someone might add video search to a shopping workspace and shape a web experience around their task.
Research output
Jacaranda: Grafting User-Editable Web Experiences Across Sites is accepted as a UIST 2026 poster and is to appear in the adjunct proceedings. The linked portfolio demonstrates design prototypes.
UIST poster authors
Xinqi Zhang, Mingyuan Zhong, Longjie Guo, Bryan Min, Yuanhao Zhang, and Kai Lukoff.