April 2024 - Present

Data Scientist

Postal (Recently Acquired by Sendoso)

• Defined and monitored business-critical KPIs to drive data-informed decision-making across the company
• Automated multi-source data workflows via API integrations in python, saving hours of tedious manual work
• Improved data accessibility across the organization by consolidating siloed sources into interactive dashboards, making critical business metrics easily accessible and reducing reliance on manual data requests
• Developed a data-driven account health scoring algorithm using key behavioral and financial indicators, empowering Customer Success to proactively reduce churn by identifying and supporting at-risk clients
• Revamped churn forecast model, increasing accuracy by 7% via improved feature selection and engineering
• Led the redevelopment of DBT data transformations in Snowflake to support a CSM platform migration

March 2023 - January 2024

Data Scientist

Pebble

• Implemented and iteratively refined a risk-based pricing model in Python to estimate group healthcare spend, achieving a 6% reduction in MAE and enabling the development of more affordable HRA-driven health plans
• Dramatically improved run-time efficiency of risk model code by 80%, enabling faster decision-making
• Built an interactive dashboard in HEX to track monthly HRA spend, providing real-time insights into spending patterns
• Collaborated closely with the sales team and clients to develop tailored health plans, aligning with unique client needs
• Enhanced health plan proposal efficiency through the implementation of an automated proposal generation system, achieving a notable 60% reduction in proposal creation time

March 2022 - June 2022

Data Science Intern

Netflix

• Web-scraped external data on over 3.4 million apps from both the App Store and Google Play Store
• Utilized AWS SageMaker to store over 1.5 GB of transformed data in Amazon S3 for secure and instant access
• Trained a linear regression model to help identify important trends in the mobile gaming market
• Cleaned and derived a table of aggregate variables from 7.6 million rows of internal customer viewing data using SQL to better characterize subscriber behavior and tendencies
• Applied PCA and k-means to form distinct customer segments
• Combined external and internal findings to formulate actionable business recommendations to grow their mobile gaming service

January 2022 - March 2022

Data Science Intern

AT&T

• Utilized PySpark and SQL to clean, wrangle, and analyze hundreds of GB worth of customer data
• Developed models, examined trends, and extracted key insights to map the typical customer journey and derive a customer-centric segmentation strategy
• Derived dozens of descriptive aggregate variables to be used for segmentation
• Segmented over 250 million customers into 7 distinct segments using K-modes, resulting in the ability to derive more personalized retention strategies
• Produced meaningful, uncomplicated visuals via Tableau and Contour to express the behavior of different customers throughout their journey
• Formulated actionable recommendations for improving customer retention at all stages of the customer journey, specifically targeting “at-risk” customers