Software Engineer

I evaluate AI-generated audio outputs against structured rubrics across four quality dimensions and provide written feedback to identify recurring model failure modes.

I document defect patterns across evaluation batches to improve labeling consistency and reduce ambiguity in downstream training data.

I built and evaluated Python classification models to predict early lung-cancer risk using National Lung Screening Trial data for more than 1,000 patients. Random Forest achieved 98% accuracy and Decision Tree achieved 95% accuracy.

I engineered the data pipeline for cleaning, feature selection, train/test evaluation, and metric scoring, including accuracy, sensitivity, specificity, precision, and F1. I also co-authored technical documentation with a six-person research team on model trade-offs for clinical decision support.

I built four Tableau and Power BI dashboards that transformed raw datasets into decision-ready insights for cross-functional stakeholders.

I iterated on report designs using stakeholder feedback to improve the clarity and communication of key findings.

Expected graduation in July 2029. Relevant coursework includes Data Structures and Algorithms, Object-Oriented Programming, and Calculus I-II. Completed Harvard CS50x and CS50P online certifications covering C, Python, SQL, JavaScript, and web development.