Bio

I'm Roslyn — a data scientist and researcher passionate about the human side of AI. I graduated from Washington State University with training spanning social psychology, neuroscience, programming, and advanced statistics.

My work explores how AI systems are built, evaluated, and deployed responsibly:

News

Projects

SQL · DuckDB

Customer Churn & Retention Analysis

SQL-first analysis of 64K customers to identify churn drivers across subscription tiers, contract types, and behavioral signals.

Python · ML

Cervical Cancer Risk Prediction

End-to-end classification pipeline with feature engineering, cross-validation, and ROC-AUC evaluation on a public health dataset.

Statistics · R

US Movie Revenue Case Study

Regression models on 2006 U.S. films — opening theaters and sequel status drive opening revenue; critic scores explain ~80% of total domestic variation.

ML · LLM

Phenomics LLM for Raspberry Breeding

Sensors · ML

Gerontechnology — Cognitive Decline Detection

Research

  • The neurobiological impacts of depression and anxiety on reward processing. [LINK]
  • Machine Learning

    Coding Neural Networks from scratch

  • Contact

    You can also get in touch with me via email (willoughbyroslyn@gmail.com)