Predictive Modeler

Predictive Modeler — Vanguard Analytics Corp

Chicago, IL | August 2023 – Present

• Developed and deployed an XGBoost customer churn model using Python and AWS SageMaker, reducing annual customer attrition by 14% and saving $450,000 in retention costs.
• Streamlined data processing pipelines utilizing PySpark and SQL, reducing feature engineering compute costs by 22% and cutting daily ETL processing time from 5 hours to 3.5 hours.
• Collaborated with product teams to design a real-time recommendations engine using TensorFlow, which boosted click-through rates by 18% across 4 primary digital channels.
• Designed and ran A/B tests to validate 6 predictive models, improving conversion estimation accuracy by 15% and minimizing false positive classification rates by 9%.

Associate Predictive Modeler — Apex Data Solutions

Evanston, IL | June 2020 – July 2023

• Built and validated credit risk scoring models using R and SQL, which reduced loan default rates by 12% while maintaining compliance with credit underwriting policies.
• Automated model performance monitoring dashboards using Tableau and Python, reducing manual reporting time by 15 hours per week for a team of 8 analytics professionals.
• Implemented principal component analysis (PCA) to compress 120+ feature variables, reducing model training time by 30% and improving overall prediction latency by 25%.
• Optimized hyperparameter tuning for random forest models using Scikit-learn, increasing target prediction precision by 11% and saving $40,000 in infrastructure expenses.

Predictive modeling, Statistical analysis, Machine learning, Feature engineering, A/B testing, Data pipelines, Python, R, SQL, PySpark, Scikit-learn, XGBoost, TensorFlow, Tableau, AWS SageMaker, Git, Docker

E-Commerce Demand Forecaster

Role: Lead Predictive Modeler

Tools: Python, XGBoost, PySpark, AWS

Designed and implemented a demand forecasting model for 15,000 SKUs, reducing inventory holding costs by 18% and improving replenishment accuracy by 22%.

Ad-Click Conversion Engine

Role: Predictive Modeler

Tools: TensorFlow, Scikit-learn, Docker, SQL

Created a real-time ad click-through prediction engine handling 2.5M daily requests, which increased advertising revenue by 14% within 6 months.

• Certified Analytics Professional (CAP) (2022)
• Google Cloud Certified Professional Data Engineer (2023)
• SAS Certified Predictive Modeler (2021)

• Languages: English (Native), Spanish (Conversational)
• Volunteer Work: Data analysis mentor for local non-profit youth organization (2022-present)
• Availability: 2 weeks notice