Trey Schulman

Experience

1. Data Scientist

   Deloitte

   Sep 2021 — Present

   Agile developer on a startup-environment product to detect healthcare fraud, waste, abuse.+ Independently developed rule-based indicators and ML models to assess different types of healthcare fraud risk implemented using PySpark/PyTorch. This process includes driving the ideation, EDA, design, development, validation, testing, and documentation for client delivery.+ Collaborated closely with client, data visualization, and ETL teams to ensure cohesive product and frequently met with client to explain mathematical concepts and new data analytics output.+ Performed data field mapping and data schema design for new clients. Responsible for troubleshooting new client data output and answering client questions and concerns.+ Led application of risk engine towards new usages and responsible for new risk scoring algorithm and fraud indicators in spaces outside healthcare.+ Built multi-label classification neural network in PyTorch for linking healthcare claims & prescriptions data.+ Researched, developed, and implemented stability metric to measure model output drift, involving complex statistical methods (Jaccard index, Kernel density estimation) and creation of related Azure pipeline configured to run on EMR cluster.+ Created new costs-savings local testing environment for whole product team involving management of Spark dependencies, environment variables, and python libraries in virtual environment to match EC2 agents in production.

Education

• University of Virginia

  Bachelor's degree, Computer Science

  2017 — 2021

• University of Illinois Urbana-Champaign

  Master's degree, Computer Science

  2024 — 2026