Department

University of Tennessee at Chattanooga. Dept. of Psychology

Publisher

University of Tennessee at Chattanooga

Place of Publication

Chattanooga (Tenn.)

Abstract

As members of the FOCUS Team within HR Science at Appalachian State University, we contributed to developing and refining an AI-powered chatbot, named Eddie, designed to enhance Eugene, a career exploration platform for Appalachian State University undergraduate psychology students. This effort was driven by findings from a Eugene Assessment survey, which revealed that over 40% of respondents believed the platform's structure and instructions could be improved. In alignment with FOCUS' mission to address psychology student underemployment by connecting students to career and educational opportunities, our team worked to improve the user experience through AI technology. Student feedback specifically identified challenges related to navigation, understanding results, and interpreting instructional materials, creating an opportunity for enhanced support. Our team was responsible for chatbot development, guardrail testing, error identification, response refinement, documentation creation, and collaboration with technical stakeholders. Using Microsoft Copilot Studio and GPT-4.1, we worked to improve chatbot performance, strengthen safety measures, develop knowledge resources, and support future implementation within the Eugene platform. One major initiative focused on strengthening Eddie's safety and ethical guardrails. Through multiple rounds of structured testing, we evaluated how the chatbot responded to inappropriate, harmful, or off-topic prompts and identified areas for improvement. The team developed standardized refusal messages, crisis response protocols, and keyword-triggered resource referrals to ensure students received appropriate guidance while maintaining the chatbot's intended purpose. These efforts improved consistency, reliability, and user safety across a wide range of interactions. A second initiative focused on enhancing the chatbot's knowledge base and functionality. We created and refined reference documents, tested response accuracy, analyzed system errors, and evaluated how effectively the chatbot utilized information from Eugene and other university resources. Through iterative testing and continuous refinement, we improved the quality, accuracy, and usefulness of chatbot responses while creating documentation to support future development. This experience demonstrated the value of combining psychological principles with emerging technologies to address real-world challenges. Through the project, we gained experience in user-centered design, interdisciplinary collaboration, stakeholder engagement, and iterative improvement. Most importantly, we learned how psychology students can contribute to the ethical development of AI systems by applying their understanding of human behavior, communication, and user needs to create technologies that support student success and career exploration.

Date

November 2026

Subject

Industrial and organizational psychology

Document Type

posters

Language

English

Rights

http://rightsstatements.org/vocab/InC/1.0/

License

http://creativecommons.org/licenses/by/4.0/

RCIO 2026 Chatbot Poster.pdf (590 kB)
Poster PDF

Included in

Psychology Commons

Share

COinS
 
Nov 7th, 12:45 PM Nov 7th, 1:45 PM

Developing a Chatbot: Enhancing Psychology Career Exploration Through AI

As members of the FOCUS Team within HR Science at Appalachian State University, we contributed to developing and refining an AI-powered chatbot, named Eddie, designed to enhance Eugene, a career exploration platform for Appalachian State University undergraduate psychology students. This effort was driven by findings from a Eugene Assessment survey, which revealed that over 40% of respondents believed the platform's structure and instructions could be improved. In alignment with FOCUS' mission to address psychology student underemployment by connecting students to career and educational opportunities, our team worked to improve the user experience through AI technology. Student feedback specifically identified challenges related to navigation, understanding results, and interpreting instructional materials, creating an opportunity for enhanced support. Our team was responsible for chatbot development, guardrail testing, error identification, response refinement, documentation creation, and collaboration with technical stakeholders. Using Microsoft Copilot Studio and GPT-4.1, we worked to improve chatbot performance, strengthen safety measures, develop knowledge resources, and support future implementation within the Eugene platform. One major initiative focused on strengthening Eddie's safety and ethical guardrails. Through multiple rounds of structured testing, we evaluated how the chatbot responded to inappropriate, harmful, or off-topic prompts and identified areas for improvement. The team developed standardized refusal messages, crisis response protocols, and keyword-triggered resource referrals to ensure students received appropriate guidance while maintaining the chatbot's intended purpose. These efforts improved consistency, reliability, and user safety across a wide range of interactions. A second initiative focused on enhancing the chatbot's knowledge base and functionality. We created and refined reference documents, tested response accuracy, analyzed system errors, and evaluated how effectively the chatbot utilized information from Eugene and other university resources. Through iterative testing and continuous refinement, we improved the quality, accuracy, and usefulness of chatbot responses while creating documentation to support future development. This experience demonstrated the value of combining psychological principles with emerging technologies to address real-world challenges. Through the project, we gained experience in user-centered design, interdisciplinary collaboration, stakeholder engagement, and iterative improvement. Most importantly, we learned how psychology students can contribute to the ethical development of AI systems by applying their understanding of human behavior, communication, and user needs to create technologies that support student success and career exploration.