Curi AI Mentoring Platform

Curi was developed as a conceptual prototype to examine how technology can be used to reduce the opportunity divide that often stems from the digital divide. Our research began with a deep exploration into how access to devices, internet bandwidth, and educational resources influences the ability of individuals to pursue career advancement.Through a combination of quantitative surveys and qualitative interviews, we found that many users who would benefit most from mentorship lack access to personal laptops or consistent internet. Text-based and asynchronous mentorship emerged as the most accessible option for these users.These findings informed the prototype’s core functionalities. Curi was designed to offer low-bandwidth, mobile-friendly, text-first interactions that are accessible in environments with limited connectivity. We are now preparing for implementation within the next year based on these research insights.This work would not have been possible without the dedication and talent of our global team, including Lauren Choi, Seongbin Lee, Hyunjong Joo, Giah Kim, Hyobin Cho, and Yeji Shim.

As a conceptual prototype, one of the most complex challenges we faced was designing an AI system that could simulate the empathy and authenticity of human mentorship. While AI is powerful in delivering information, mentorship involves nuance, reassurance, and context.We created detailed prompt structures and sample dialogues modeled on real mentor interactions. These helped shape how the AI mentor should respond based on different levels of user experience and career readiness. Our team was also attentive to avoiding bias and ensuring cultural sensitivity across different contexts.As we move toward implementation in the coming year, we plan to use feedback from real users to improve the personalization and emotional intelligence of the AI system.

Our global team brought together a variety of lived experiences that played a central role in shaping the conceptual prototype. Designing Curi to be both accessible and relevant across regions required us to account for differences in cultural norms, communication styles, and mentorship expectations.Seongbin Lee, based in Singapore and fluent in English, Korean, and Chinese, brought insight from his life in South Korea, China, and Japan. His contributions in software and user research helped us test the assumptions we might have made within a single regional context.This international collaboration allowed us to build a more inclusive foundation for the platform. As we move toward implementation, we will continue to incorporate perspectives from our global team to ensure Curi’s accessibility and cultural relevance.