Developed new production ready credit decisioning models for all loan product lines including installment, line of credit, and payday products for both collateralized and and uncollateralized loans using XGBoost and scikit-learn.
Reduced default rate for installment and line of credit products by over 20% and increased annualized net cash by $2.7 million.
Created a robust data pipeline and feature store with Python, FastAPI and SQL Server to make the feature creation process scalable and reliable, and to automate feature computation, backfilling, and logging.
Worked closely with internal stakeholders on creating regression models to better understand how to optimize returning customer underwriting.
Data Scientist
Early Warning
07.2021 - 03.2023
Used Spark, SQL (HiveQL), and Python for exploratory data analysis and visualization to search for insights from Zelle transaction and customer data
Predictive modeling using Python, scikit-learn ML models & Data
Utilized advanced querying, visualization and analytics tools to analyze and process complex data sets.
Built features and trained highly accurate XGBoost classifier to predict fraudulent Zelle payment transactions
End to end feature engineering - brainstorm, create, validate, down-select - to facilitate proof of concepts and rapid prototyping of new ML data products
Built internal fraud analytics packages in R with functions to quickly link fraud threats to specific attributes
Communicate impact of changes to strategies and fraud losses to key stakeholders
Data Analyst
Associations International
07.2020 - 06.2021
Developed predictive models for member lifetime value and member retention across several cohort structures using python and Liftetimes library
Analyzed, interpreted, and presented model output and findings to organizational stakeholders, including executive leadership
Collaborated with marketing team to develop A/B testing plans to assess different variations of website member profile functionality leading to a 15% increase in conversion rate
Developed dashboards to provide insights and visualization into member and channel performance based on existing and new KPIs, channel projections, and historical performance
QA Engineer
Kuhl Media, LLC
08.2018 - 09.2020
Implemented web and API testing frameworks
Developed strategies for comprehensive test plan automation within agile software development lifecycle integrating with cross-functional teams
Used SQL to create, modify, and execute complex queries for testing backend data integrity testing, business logic, and API flow
QA Engineer/Analyst
Experis
10.2014 - 11.2015
Excelled in several long term contract positions through this agency
VML / QA Engineer - Part of an agile team implementing the redesign and architecture overhaul of Sprint’s website and mobile apps. Set up automated testing for mobile apps. Led team developing, coding, and implementing API test suite.
Sungevity / QA Analyst - Added to existing testing codebase for several large new APIs that serviced both Sungevity’s customer facing assets and internal analytics, sales, and mapping clients. Increased coverage on existing codebase from 35% to 85%.
VML / QA Analyst - Set up mobile app testing for several large clients using Appium and Sauce Labs including Medhost, Dell Computer, and Quick Trip. Removed 60% of existing manual testing.
Cerner / Solution Designer
07.2013 - 10.2014
Defined and tracked key metrics for several projects in the population health space
Collaboratively analyzed business requirements with multiple sets of stakeholders including architects, strategists, and project managers to provide execution planning
Prioritized and supported backlog, issue triage, and wrote test cases for enhancements and defects