- Managed creation of a series of Product Hierarchy Preference Models that power personalized email and site experiences delivering on customer personalization capabilities.- Designed and implemented a centralized feature store for all data science models minimizing project startup time and data quality errors due to lack of data governance.

- Created Forecasting Model and Forecasting Report for monthly, quarterly, and yearly customer and financial metrics enabling Senior Leadership Team across the portfolio of brands to engage in both near and long-term planning enabling reflection on brand strategy.- Collaborated with a cross-functional team to mature returns data and produce Returns Dashboard delivering elevation of product attributes attributing to returns within Banana Republic brand.- Partnered with multiple teams to produce Customer Segmentation Dashboard powered by Segmentation Model driving identification and tracking of target customer segments for all brands in portfolio.

- Assisted in increasing online sales by 61% through ETL of online customer data using Databricks (SQL) and Adobe Data Warehouse/ Adobe API (Python).- Mentored and trained data engineers and data scientists around automation and productionalization of data science models, analytical reporting, and query optimization.- Create reports and 10+ models to improve delivery of goods and services to customers- Mentor and train junior engineers and 20+ data scientists concerning automation and productionalization

- Built and presented data science model productionalization system to COO and SVP of Data and Analytics. Project resulted in savings of 8 weeks and $80,000 per data science model implementation through creation of model syndication pipeline (SQL) and data science productionalization helper functions (Python and R).

- Created source of truth dataset for longitudinal customer data.

- Goal of capstone was to present a plan to theoretically reduce the wait time of the Uber rider population in NYC. - Improved upon baseline accuracy by 56%- Used a variety of machine learning techniques (primary model was a deep neural-net) to successfully complete a real-time spatiotemporal forecasting model for the Uber rider population in NYC. - Compared the results of Deep-Learning , Linear Regression and Random Forest models to gain a better understanding of the pros and cons of each model in a real-world application.- Transformed forecasted predictions into longitude and latitude coordinates and plotted them in Google Maps for stakeholder reference- Formulated rollout of real-time forecasting tool to direct drivers to future rider populations to effectively reduce average wait time of Uber rider and increase average number of rides by driver.

- Achieved Accounting department goal of decreasing month-end closing efforts to sub-week by identifying and automating over 2000 work-hours of data processes.