Gilbert, Arizona, United States
Computer Science and Mathematics major at Duke University
Developed a full‑stack platform that automates discovery, extraction, and AI‑driven interpretation of U.S. county solar‑ordinance and sentiment data. The system combines a DuckDuckGo/Google search wrapper, a resilient PDF/DOCX/OCR scraping layer, and classifiers that score both ordinance severity and community sentiment using cosine similarity metrics and LLMs. Results are persisted in a normalized PostgreSQL schema and surfaced through a Flask‑based web map, enabling analysts to trigger or resume county‑level pipelines on demand. Beyond this core platform, I partnered with the permitting team to streamline approvals via Power Automate and leveraged SQL + Databricks to prototype machine‑learning models predicting when projects would exceed budget. Collectively, these initiatives demonstrate end‑to‑end data engineering, applied machine learning, and user‑facing visualization skills suited for software engineering or data‑science roles.
Organized and migrated cervical cancer imaging data from a disorganized folder system into a MongoDB database, using Python scripts to automate processes while ensuring data integrity and maintaining patient privacy.