CloChat is a user interface that enables users to customize and interact with agent personas in large language models (LLMs). The system allows users to define various aspects of an agent's persona, such as personality, communication style, and visual representation, to create a more personalized conversational experience. Through an exploratory study comparing CloChat with ChatGPT, the research found that users formed stronger emotional bonds with customized agents, engaged in more dynamic dialogues, and showed greater interest in maintaining interactions. The study also revealed that agent personas customized with CloChat significantly enhanced user engagement, trust, and emotional connection compared to generic ChatGPT. Additionally, the personas and dialogues in the study were derived from the main study, which explored how users customize and interact with agent personas in LLMs. The research contributes to the understanding of how users customize, interact with, and experience personas in LLMs, and highlights the importance of personalization in conversational agents. The study also identified ethical considerations in persona customization, such as the potential for biased representations and privacy concerns. Based on these findings, the research proposes design implications for future conversational systems using LLMs, emphasizing the need for personalized and user-centered approaches. The study's results suggest that personalized agent personas can significantly enhance user experience, making interactions more engaging and relevant to individual users.CloChat is a user interface that enables users to customize and interact with agent personas in large language models (LLMs). The system allows users to define various aspects of an agent's persona, such as personality, communication style, and visual representation, to create a more personalized conversational experience. Through an exploratory study comparing CloChat with ChatGPT, the research found that users formed stronger emotional bonds with customized agents, engaged in more dynamic dialogues, and showed greater interest in maintaining interactions. The study also revealed that agent personas customized with CloChat significantly enhanced user engagement, trust, and emotional connection compared to generic ChatGPT. Additionally, the personas and dialogues in the study were derived from the main study, which explored how users customize and interact with agent personas in LLMs. The research contributes to the understanding of how users customize, interact with, and experience personas in LLMs, and highlights the importance of personalization in conversational agents. The study also identified ethical considerations in persona customization, such as the potential for biased representations and privacy concerns. Based on these findings, the research proposes design implications for future conversational systems using LLMs, emphasizing the need for personalized and user-centered approaches. The study's results suggest that personalized agent personas can significantly enhance user experience, making interactions more engaging and relevant to individual users.