Data Scientist | Math PhD | 25 Years Solving High-Stakes Analytical Challenges

Statistical rigor + modern ML for anomaly detection, risk assessment, and impact measurement. Currently: fraud detection research, mathematical consulting, and open-source analytics projects.

• PhD mathematician specializing in custom analytical solutions that reveal insights standard approaches miss
• 21 federal awards for analytical innovation and cross-functional collaboration
• Specialized in: Anomaly detection, risk quantification, impact measurement, predictive modeling, A/B testing, Monte Carlo simulation, constrained optimization, quantifying uncertainty
• Full analytical lifecycle: Stakeholder requirements → data challenges → analytical solutions → deployment

What I bring:

• Deep statistical foundation enabling custom methods beyond typical data science
• Production-oriented thinking optimizing for business value
• Clear communication translating complexity for diverse decision-makers

Experience:

• 24 years vehicle safety analysis and risk assessment (NHTSA)
• 3 years national economic indicators (U.S. Census Bureau)
• 6 years college teaching
• fraud detection R&D portfolio (8+ open-source projects)
• statistical consulting for an automaker
• mathematical consulting for AI research

• squeezing-more-info-out-of-fraud-data-w-statistics – Reveals fraud signals ML alone misses
• how-bad-could-this-emergent-fraud-be – 30% accuracy improvement in estimating potential fraud losses
• medical-upcoding-save-tens-of-thousands-per-yr – Provider-level clustering detecting systematic patterns

• Fraud-FPR trade-off analysis, investigative staffing optimization, cost-sensitive model comparison

• my-posts-on-fraud-detection – ML model benchmarking with custom CardPrecision@k/CardRecall@k metrics
• db_fraud_detection – Databricks dashboard for transaction flagging and drift monitoring
• hf_fraud_detection_space – Hugging Face/Streamlit API for interactive predictions
• ieee-fraud-detection-dbt – Snowflake/dbt pipeline for data ingestion and validation

• The Math Behind Fraud Detection with Logistic Regression – Mathematical foundations of optimization and hyperparameter tuning
• Fraud Detection: Same & Different – How ML for fraud differs from other domains

Full portfolio: dglassbrenner1.github.io

LinkedIn: Donna Glassbrenner, Ph.D.
Portfolio: dglassbrenner1.github.io
Consulting: Analysis Insights, LLC - Available for statistical consulting and analytical projects