David Elgas

Data Analysis and Machine Learning Projects

Predictive Analytics and Explainable AI In this project, I developed an AI model to identify and quantify the key factors contributing to customer churn in a subscription-based service. The model not only predicts churn with usable accuracy and precision, but also provides insightful interpretations of the underlying causes. The results shown were implemented recently in a subscription-based business and reduced churn by > 15%. Notebook
Conversation Modeling This project aims is part of a large effort to leverage Generative AI to create a "virtual mechanic." Here, LDA is used to find unique topics for proper summarization, clustering and encoding. Notebook
Large Language Models with RAG This project aims to leverage Generative AI to create a "virtual mechanic." By utilizing a corpus of text gathered from a car forum, this AI-driven tool will be capable of understanding unstructured questions and providing relevant answers. Report
Statistical Analysis of Homelessness Homelessness represents a persistent and growing challenge in the United States with the magnitude of homelessness particularly evident in Los Angeles, California . This study was undertaken to look for possible answers that may ultimately help this underserved community. Notebook Report
Project Sandbox A collection of experimental projects and works in progress. This sandbox environment showcases various technical implementations, proof-of-concepts, and innovative ideas across machine learning, data analysis, and AI applications. These projects represent ongoing exploration and development in emerging technologies. Project Directory