I am currently working with Poppins On-Demand Childcare to help them build a new website and customer portal. The goal of this project is to make it easier for customers to sign up for childcare services, manage their accounts, and communicate with the company. I am using a variety of technologies including Wix, Velo JavaScript, and payment tools to build the website and portal. I am also implementing a custom content management system to allow the company to easily update their website content without needing to know how to code.

The modern media environment is flooded with AI generated content. The need to distinguish between real and AI generated images has become increasingly relevant. The goal of this research was to leverage machine learning strategies to classify images as real or AI generated.

The research was in collaberation with the UVU Master's of Computer Science Program and published through IEEE and presented at the 2025 UVU i-ETC conference in a paper titled "Machine Learning Classification of AI and Real Images". Authors include myself, Joshua Hall, and my colleagues at the UVU MCS Dr. Jingpeng Tang and Dr. Ting Cao.

Testing software is one of the most difficult parts of developement and often requires very large teams of prefessionals. ASTA makes the testing process much easier by allowing anyone to use plain English to specify application requirements and flows.

Winvoice is a team project React App currently in developement using MERN technologies. This app's goal is to help simplify invoice storage, views, creation, and editing. So far, we have a working REST API that generates unique id's for invoices and sets of invoices. We have responsive React tables using different node libraries. My contribution to this project was the bulk upload process. This process allows a user to upload CSV files with one or many invoices and the system can recognize key field names for validation and integration into the databases. I also set up page routing using React Routers. We made this project totally from scratch.

During my time a Videru, I helped develope their targeted advertising system. This system would use video input and a Machine Learning model to determine what advertisements a customer would most likely be interested in. I came in and set up the company's version control using GitHub's large storage service extension. I also made it possible to upload different formats of advertisements including many video and image file types.

This project was a semester web developement project. Tools used include NodeJS, Bootstrap, and Express. I designed a local storage service API for it to consume. Views, models, and controllers were also added with template HTML generated through EJS. The final product consumed an SQL database using another REST API that I designed.

This stock price prediction model was designed by me and my back-end team at a summer internship at UVU. At first, the machine learning model was not great at preditions (guessing the correct price trend with about 55% accuracy). To help, I made an alternate model using GRU instead of LSTM to prevent some over-fitting issues. I also made a way to visualize the predictions on a graph as well as a script to automate the training process.