As part of my Master’s thesis, I conducted research on the usage of UI agents in gestural interfaces. The research focused on providing visual affordances through UI agents for large-screen displays. User engagement is critical for many public information systems or large-screen displays. The gestures used for the interface need to be easy to understand. Previous research has shown that users need feedback for understanding natural gesture interaction. To achieve the goals of the thesis, I built three prototypes in an iterative model. These explore different ways of using UI agents in providing visual cues. A focus group and two user studies were conducted to test the prototypes. Prototypes were evaluated based on initial user engagement and system usability on the main interaction phases. Results of the user studies show that using UI agents as visual affordances is more engaging and results in fewer errors during gesture interaction. The success of the UI agent depends on its relationship with the interface design. Overall, UI agents are effective in giving users feedback in order to help them understand the interface. These findings are important for designing public information systems where user engagement is required.