Projects
The INSIGHT Lab is actively hiring graduate students for these projects. Please visit the join page to learn more. Topics beyond the scope of these projects are also welcome.
Human-AI Collaboration in Data-Driven Decision Making
A crucial, time-consuming, step in data-driven decision-making is systemically analyzing data to uncover insights in the form of patterns, trends, outliers, and other anomalies (or lack thereof). This group of projects will investigate to what extent, and in what ways, AI agents can collaborate with teams of data workers to discover and prioritize actionable insights for decision-making.
Human-Centered Evaluation of AI generated Code
In collaborative team settings, individual data workers prioritize different factors when assessing insights, as do AI agents. Projects in this space will explore techniques that allow humans and AI agents to externalize and align priorities when assessing insight quality. These projects will also seek to modify AI agent behavior to better align with human priorities.
Provenance in Human-AI Collboration
As humans and AI agents work collaboratively, it becomes important to understand how decisions are made, what factors influence them, and how both human and AI inputs contribute to the outcome. This group of projects will explore mechanisms for capturing, representing, and using provenance information.