Graise Zhou, SEAS Masters in Computer and Information Technology, 2024
I was given the assignment of building a chatbot prototype while working as an AI Development Intern at Mentorz, a developing social network for mentorship. This chatbot’s main objective was to assist mentees in creating specific goals for their mentoring sessions. By analyzing a mentor’s profile information, including LinkedIn links, personal websites, areas of specialty, and introductions, this was accomplished. The GPT-3.5 Turbo model from OpenAI served as the brain of this chatbot, and I used it to create logical prompts. When the model responded to these questions, it provided attainable objectives, improving the mentees’ preparation process and enhancing the mentorship experience as a whole.
When I think back on the first part of my internship, a number of successes jump out. I was proud of myself for successfully implementing the chatbot prototype’s initial iteration through the OpenAI API, which demonstrated skill in deciphering mentor profile data and launching appropriate prompts. The GPT-3.5 Turbo model’s AI capabilities were added to this chatbot, enhancing it further and enabling mentor-based goal recommendations. My commitment to enhancing user experience is seen in the feature I introduced, which limited AI responses to 200 words in order to provide concise, useful insights. Furthermore, I developed a method to gather relevant mentor profile data from many sources using my academic expertise in database management.
There are many goals I have when I consider what my internship may entail in the future. My top focus is to improve the chatbot and increase its ability to recommend goals by adjusting performance based on user feedback. We will carefully conduct user testing, a crucial component of product development, to find any functional or UX flaws. Any AI-driven product needs a constant feedback loop, and I’m about to roll out a system to collect customer feedback to make sure the chatbot keeps improving and staying in line with user needs. Finally, documenting my development process and outlining the chatbot’s features is planned, providing a useful reference for the development team and any subsequent owners of the project.
The horizon provides various focal points that become apparent. The road to creating the perfect chatbot is long, but I see it developing its capacity to recognize and respond to complex user requests. Optimizing its performance and ensuring its agility even in situations of increased user involvement is a parallel track I’m interested in. My obsessive curiosity drives me to study more about AI and machine learning, particularly the subtleties of natural language processing, which can greatly improve the chatbot. My efforts will also include improving our chatbot utilizing MindsDB for greater SQL database synchronization and adding a speech-to-text solution for the platform using OpenAI’s Whisper API.
In conclusion, my time at Mentorz has been a mix of problems overcome with creativity and tenacity. After laying the groundwork, I’m anxious to further develop the Chatbot to make sure mentees get the most out of their mentorship relationships.
This is part of a series of posts by recipients of the 2023 GAPSA Summer Internship Funding Program that is coordinated by Penn Career Services. We’ve asked funding recipients to reflect on their summer experiences and talk about the industries in which they spent their summer. You can read the entire series here.
