• Led efforts at developing 12 new project ideas for new club members during Fall Quarter by matching interested members with ongoing and new projects; meeting with project leads weekly to check in on project status. • Engaged industry leads from SCU’s alumni network to enhance mentorship opportunities for projects.

• Collaborated with Dr. Kai Lukoff to analyze demographic bias in 100+ LLM-based chatbot conversations related to career personas using Claude and GPT-4; automated prompt generation using GPT-4 API and Claude 3.0 API.• Employed text-mining and NLP techniques (i.e. sentiment and word cloud analysis) to identify bias patterns in textual output.

• Engineered a web application using Node.js and React.js to automate visualization of 1,000+ customerinteractions with company’s cloud-based virtual machine platform using Grafana’s API.• Built a machine learning model to calculate prices for customers using company's cloud-native virtual machine services with over 95% accuracy using Linear Regression.• Developed a GPT model using PyTorch for an LLM-based chatbot to investigate customer activity on cloud-based virtual machine usage.• Created a Python script tagging specified Git commits and storing previous 10 commit messages of a given repository into a JSON file to enable automated documentation of code changes and development history.• Deployed a visualization of metrics from customer interaction with company’s cloud-based virtual machine platform using Grafana, Loki, and Promtail.

• Collaborated with Dr. Kai Lukoff to utilize developmental direction and output of 5 character chatbots and 3 general chatbots for designing a character-based therapist chatbot.• Studied articles as part of a research group on generative AI in human-computer interaction and other related fields to identify research directions for designing novel user experiences.

• Engineered a crypto-trading bot model as a key contribution to Cisco’s Bold Bets initiative to join previously untapped industry markets by integrating a robust back-tester functionality and an advanced analytics viewer.• Conducted research and analysis with the Polaris operating system team, using several systems design concepts (i.e. stateless architecture and database sharding), to understand how to scale software applications across more than 1,000 users with multiple features.

• Built an algorithm analyzing attributes of individuals’ loans from a dataset for predicting if the loans would be paid back, charged off, defaulted, or current with an accuracy score between 88% and 90%. Utilized Logistic Regression and Ridge Classification.• Created prediction models on real-world industry datasets, with accuracy scores of 80% and higher, by utilizing Natural Language Processing, Ridge Classification, and Lasso Regression.

• Learned about the Airspace Galaxy (drone tracking) UI by assisting with tracking and debugging errors. • Performed on-site testing and debugging for the Airspace Galaxy system, tracking 10 to 20 errors per day.• Worked with the iMacros and XCTest frameworks to debug errors on the Galaxy UI and program test scripts in VBA (iMacros), and Swift.