PII Safeguard is a Chrome browser extension tool designed to enhance data security by discovering, identifying, and redacting personally identifiable information (PII) across different files to protect sensitive information. By using a prototype designed in Figma that outlines the tool’s user interface and user interaction workflows, and leveraging a backend that's built with Python, the browser extension scans inputted files to detect over 40 different categories including financial, personal, and security entities. This software tool aids users in managing and securing sensitive PII, ensuring compliance with privacy standards and reducing the potential risk of data breaches.

The proliferation of AI-generated images and deepfakes on social media platforms has fueled misinformation and ethical concerns. This project addressed the need for reliable detection tools by developing an AI Image Detector Chrome extension to inform users if celebrity photos on Instagram are likely AI-generated. Transfer learning techniques were employed to fine-tune pre-trained CNN models on a dataset of real and AI-generated celebrity images. While performance analysis revealed inconsistencies, progress was made in generating AI images, training models, and developing front-end functionality. Moving forward, we aim to expand the dataset, incorporate explainable AI features, enable mobile integration, and conduct user-centric refinement. Our work strives to provide a transparent tool for identifying AI-generated content, empowering users to combat misinformation, and promoting ethical online practices.

For over 100 years, The Daily Utah Chronicle has published the student voices of the University of Utah. Many of these stories have been archived, but there are significant chunks missing. Just over 23% of the Chrony’s work has not been archived, and there are no procedures in place to keep the archive updated. I wrote a manual for them containing an in-depth analysis, a plan of action, and a concise archival process for leadership to implement. This will provide a clear path to archiving the past, and be a good compass for preserving the student voices of the future.

This project evaluated models for the preservation and redistribution of the collocated Law Library of Congress digital collection material within the American Memory Project “Evolution of the American Conservation Movement, 1850-1920” collection, which has been decommissioned, making access to this material difficult. Keeping these materials together and accessible provides users with a historical context for current conservation activities and legal proceedings. The final models selected are sub-pages highlighting the collection on the Statutes at Large digital collection Articles and Essays page, a cascading timeline StoryMap showcasing important years, and In Custodia Legis blog posts promoting the aforementioned models.

Lewis County has a rich, vibrant history which shapes the lives of all those who reside and work there in ways that are complex and multifaceted. However, firsthand accounts of historical events and their impact are in short supply and, with every passing year, those who were alive to experience those events are slipping away forever. Guided by the Lewis County Historical Museum’s director, I gathered oral history interviews from county residents on events that shaped their lives and the county itself. These interviews will be preserved in the oral history archive to inform and enlighten future generations.

Scarecrow Video holds the largest publicly available video collection in the US. Their vast foreign film section includes many irreplaceable films from less established movie industries, making it a high priority for assessment in order to care for and protect these films. This project involved an audit of about 40% of their African film section, resulting in a list of films to be added to Scarecrow’s Rental by Agreement (RBA) program, as well as the identification of possible orphan works for future digitization, and guidance for further assessment and protection of this collection.

According to research, Disability Resources for Students (DRS) at the UW lacks holistic resources to equitably aid students outside of the classroom. To address this, we designed REACH, a web-application that offers UW students easy access and connection to resources both on-campus and off, connection with like-minded students through an anonymous discussion forum, and convenient answers to common questions and issues other students with dis/abilities have experienced at UW. Our mission is to empower students with disabilities by providing an online platform where they can discover a supportive community and access essential resources, tailored to their unique needs.

Sponsored by Included, an HR analytics consulting firm, our team has developed an intelligent dashboard that tackles the inefficiencies of traditional recruitment processes. Manual resume reviews are time consuming, and assessing candidate qualifications can be daunting. Our solution is a dashboard powered by two AI algorithms. Leveraging LLMs (large language models), it seamlessly matches candidate qualifications to job post requirements. The result is a ranked list of top-notch candidates, ready for recruiters to evaluate swiftly. Our tool puts the power back in recruiters’ hands. They can confidently make final decisions, ensuring a personalized touch while drastically reducing resume review time.

We examined the use of the Dewey Decimal System in a K-8 library, with the goal of increasing circulation of nonfiction materials. We provided our sponsor with an organizational framework she will use to reorganize the nonfiction section, as well as illustrated section signage and themed book kits. Our project proved that rethinking Dewey is a worthwhile endeavor, and encouraged Open Window School students to be part of the process of reimagining their library space.