HuskyNav is an AI powered navigation app designed specifically for anyone trying to get familiar with UConns Storrs campus. It was developed to solve a common problem those unfamiliar with campus face when being introduced to Storrs: room level directions. While modern GPS navigation is a solved problem, many first years struggle with finding their classes for the first time. Large academic buildings often contain outdated signage for room and floor numbers, causing confusion and inaccessibility for countless groups of students, first-time visitors, and individuals with disabilities. HuskyNav aims to alleviate stress and provide written directions through these buildings to give everyone the best possible experience on UConn’s Storrs campus. To solve this problem, we use frontier technology in the shape of a Large Language Model (LLM) to interpret building data. Specifically, our project uses Google Gemini, one of the leading LLMs in terms of accuracy and performance of responses. These models excel at responding to a variety of inputs, which fits our purposes as our data is not just in English text, but rather floorplans PDFs. We provide these floorplans to Gemini, accompanied by a prompt, and it outputs to the user written directions from the entrance of the building all the way through to their requested room number. Using the Google Maps API, our app gives students directions from their dorms to their building, and then from the building entrance to their classroom. The end goal for our project is to make living and learning on campus simple, letting students and visitors get to where they need to go, when they need to go. We collect no Personally Identifiable Information (PII), allowing for seamless usage of the application with no login or sign up process. The UI is designed to mirror UConns colors and for simplicity, to give the user a familiar and easy to use experience. At the end of the day, we hope that HuskyNav can make our fellow Huskies lives’ easier.