Analytics engineer at CarGurus. I think most data problems are trust problems, so I spend my time on the parts that make data believable - tests, contracts, clean lineage - mostly in dbt and SQL on Snowflake. I'd rather ship a model that's correct than one that's clever.
Currently
- Building Turbo - CarGuru's internal AI platform for automated data engineering
- Teaching myself Rust via Advent of Code 2025 - partly curiosity, partly to stop treating "the fast layer" as someone else's job.
Stack
- Modeling / transform: dbt (dbt Cloud), SQL, Spark
- Warehouse: Snowflake (also Redshift, BigQuery)
- Orchestration: Airflow, AWS Glue, Fivetran
- Cloud / lang: AWS (S3, Glue, Athena, Redshift), Python - and Rust on the side
- BI: Omni, Looker, Streamlit
Featured
- dotfiles — my macOS dev environment (fish, Neovim, tmux, Ghostty, starship).
- advent-of-code-2025 — learning Rust the hard way, one puzzle at a time.
Off the clock I'm usually overengineering a hobby project, gaming, or making a latte that's better than it needs to be - with a sci-fi novel within reach.