Portfolio

1. California freeway performance modeling (dbt & SQL)

I developed and documented a data model to detect traffic bottlenecks across California's freeway network. I used complex concurrent criteria and window functions to calculate spatial and temporal speed deltas across millions of rows. I developed geospatial logic for directional freeway postmiles to maintain geographically accurate upstream and downstream calculations. The model uses an incremental materialization for scalability and cost optimization. I also built a validation model to filter for high-quality sensor data.

This work translates the "Systematic Identification of Freeway Bottlenecks" algorithm into production-ready SQL using dbt.

🔗 Bottleneck model

2. Data architecture and governance

I designed a conceptual data architecture diagram illustrating Snowflake data warehouse database layers, schemas, roles, and environments to support data governance and self-service analytics.

🔗 Architecture diagram

1. Unemployment Insurance user journey analysis

I consulted with EDD to architect and implement event tracking via Google Analytics, enabling the first-ever capture of granular user behavior data on the UIO platform. I then performed a deep-dive analysis of this data, visualizing the results with Python. The insights from my analysis were cited in a report solicited by Governor Newsom, directly informing the decision to overhaul digital forms with intuitive, guided interactions. This work successfully transitioned the department toward a data-driven, iterative approach to evaluating and improving the claimant experience.

🔗 Governor's report (citation p. 46-50)

🔗 Python data analysis

1. Regular expressions in Python & PySpark

A technical guide breaking down regex concepts with real-world Python and PySpark examples, written for practitioners.

🔗 Read the article