Eva User

I’m a Data Science intern building end-to-end predictive models on real-world finance, public safety, and environmental datasets. I turn raw data into decision-ready insights, combining rigorous validation with clear stakeholder communication.

In my internship at Spinnaker Analytics, I’ve built ML pipelines covering EDA, feature engineering, model selection, and evaluation across multiple projects. I deliver stakeholder-ready reporting decks (PowerPoint) and apply train/test isolation, cross-validation, and leakage checks to keep results production-reliable.

My work includes a 3-class credit risk classification model on 100K+ customer records, where a Random Forest tuned with GridSearchCV reached 81% accuracy and 0.81 weighted F1-score. I also identify key drivers using Cramér’s V and ANOVA F-scores to support interpretable business decisions.

I’ve built forecasting and regression models too—predicting NYC civil-service salary ranges with tuned XGBoost (R² of 0.913 for minimum and 0.951 for maximum) and producing a 2022 rainfall forecast from 121 years of historical data using leakage-safe TimeSeriesSplit. I’m now seeking a Data Analyst/Financial Analyst role to apply my quantitative toolkit in a business-critical setting.