This project involves sensitive government research. Specific methodologies, data, and interface designs are confidential. The following overview focuses on my contributions to the research tooling and system architecture.
Mastopia is a research initiative focused on measuring AI trust and creating robust frameworks for assessing trust in dynamic and adapting AI systems.
The goal was to build general-purpose tools that allow researchers to understand how users interact with and trust evolving AI agents in real-time, high-stakes environments.
Built intuitive interfaces and refined user flows to facilitate natural interaction during controlled experiments.
Implemented robust event capture mechanisms to track user decisions and interaction patterns with high precision.
Assisted in structuring and preparing behavioral data, ensuring it was ready for downstream analysis and modeling.
The research findings and full system details will be available upon the publication of the official paper. A link will be provided here once released.