Anna Cavanaugh

Web-Based Image Editing Platform

• Built a full-stack web application that allows users to upload and edit images with real-time filters.
• Developed a FastAPI backend to receive, process, and apply filters (e.g., grayscale, blur) to uploaded images
• Integrated Firebase Authentication to enable secure Google-based login and user session management.
• Stored processed images in Google Cloud Storage and logged metadata in Cloud Firestore.
• Deployed backend using Cloud Run with containerized Docker images managed via Artifact Registry and CI/CD pipelines.
• Designed a responsive single-page frontend (HTML, CSS, JS) that displays uploaded images and supports interactive modals for editing.

Predictive Modeling of Paint Color Preferences 

• Conducted exploratory data analysis and developed predictive models using a custom dataset of paint color attributes (RGB, Hue, Saturation, Lightness).
• Built and evaluated multiple linear and logistic regression models to predict transformed continuous and binary outcomes.
• Applied Bayesian linear and generalized linear modeling techniques using Laplace Approximation and to quantify parameter uncertainty.
• Tuned and compared model performance using cross-validation and resampling strategies, selecting top models based on RMSE and classification accuracy.
• Visualized data trends, model predictions, and coefficient uncertainty to inform model selection and feature importance.

Skincare Website(She Innovates Hackathon)

• Collaborated with a team to design and develop a skincare ingredient comparison website, leveraging HTML, CSS, and JavaScript.
• Utilized React to enhance efficiency and user experience, ensuring seamless functionality., Worked closely with teammates to brainstorm features, optimize design, and troubleshoot issues, fostering a dynamic and innovative development process.

Work Eligibility: Eligible to work in the United States with no restrictions