Responsibilities include:

• Conducted research and methodology design for the Census Coverage Estimation program, leading data-driven projects within the Decennial Statistical Studies Division that applied statistical and machine learning techniques to evaluate and improve the quality and operational efficiency of the decennial census.
• Researched small-area population and living quarters estimation methods for 2030 Coverage Estimation, leveraging statistical models with population-level administrative records. Designed and implemented anomaly-detection and quality control frameworks.
• Advanced intercensal population estimation through the Continuous Count Study by integrating Census, commercial, and government-wide datasets. Applied log-linear and latent class modeling for characteristic imputation when implementing multiple-systems estimation. Developed dual-system estimation methodologies using the American Community Survey and administrative records.
• Developed machine-learning models for the 2020 Post-Enumeration Survey (PES) creating large-scale feature selection and imputation pipelines.

Responsibilities include:

• Partnered cross-functionally with Product, Risk Management, and Compliance to translate statistical modeling and data analysis into KPI and OKR insights through automated reporting and dashboard solutions.
• Delivered executive-level summaries to senior leadership, developing flexible reporting solutions to drive strategic decision-making and monitor performance indicators.
• Designed time series ARIMAX forecasting models leveraging public economic data to predict quarterly loan origination volume, optimizing workforce allocation and reducing operational costs.
• Utilized natural language processing and text mining to analyze unstructured mortgage process documentation, uncovering operational bottlenecks and driving data-informed workflow optimizations that reduced loan closing times.