Developing a computer vision pipeline using OpenCV and Haar cascades to automate hydrophobicity analysis on biomedical surfaces. The system detects fluid contact angles and classifies SNAT molecule solubility across various catheter and ECMO tubing materials. Also designing a front-end interface to streamline experimental data logging and testing workflows for researchers. The work is contributing to a forthcoming peer-reviewed publication and aims to improve accuracy and efficiency in previously manual analysis.

Worked on a project in collaboration with American Airlines to model baggage arrival curves at major airport hubs. Designed and trained machine learning models to forecast bag scan distributions using MLPClassifier and Random Forests, with an emphasis on solving class imbalance and generalization challenges in real-world scan data. Used Snowflake SQL for large-scale data extraction and exploratory analysis, uncovering operational patterns in flight and baggage timelines. The models were validated in collaboration with AA engineers and deployed to improve on-time performance and resource allocation at hubs like Phoenix and Dallas.

Software development and planning for key backend components of an AI-powered real estate platform aimed at automating document generation, property analysis, and transaction support. Leveraged NLP and machine learning to streamline the process. Collaborated directly with the CEO to create a mockup of the front end of the app, aligning with desired components revealed through customer discovery.

Co-founded a tech-driven advertisement startup, revolutionizing residential advertising by bringing cost efficient, streamlined solutions to local businesses. Proudly awarded the Zell Lurie Institute Dare to Dream Grant, empowering communities and local businesses through hyperfocused local advertising campaigns.