Background
University of Virginia, Class of 2026
Project
Generative Artificial Intelligence (AI) in the Workplace
Project Overview
Avery Wilkerson and I were tasked with creating a way to quantitatively test an AI chatbot before it was deployed on the Foundation’s website. We wrote Python script to automatically test the bot and return an accuracy score from 0–100 that was based on the bot’s responses.
Project Outcomes
We determined that ChatGPT 4-o is the best large language model to use based on its consistency of responses. Additionally, we revised the bot’s instruction prompts to output more accurate answers.
What new skills or knowledge did you gain during your internship?
Although I’ve worked with Python and new Python packages, I learned about retrieval-augmented generation (RAG) models* and how they are created.
What was the most important learning point of your internship experience?
Learning to communicate with my mentors.
Has your internship affected your career plans? If so, how?
This internship has provided me skills that I can hopefully use in future internships. Working with AI has allowed me to stay up to date with current trends in tech.
*Retrieval-augmented generation (RAG) is the process of optimizing the output of a large language model.