A nonprofit's journey: we walked so dbt could run
A small nonprofit data team spent years patching a fragile SQL warehouse, then found that the hardest work they'd already done carried straight into their dbt rebuild.
Overview
Planned Parenthood Federation of America's basic SQL warehouse reached its limits. Covers the transition from a monolithic, waterfall SQL setup to dbt, lessons learned, and how dbt's modularity is transforming their data analysis.
What you'll take away
- 01
PPFA's first attempt, seven standalone SQL scripts chained in a waterfall, had no code reuse and an all-or-nothing failure mode: one broken script meant no warehouse for the day.
- 02
Almost none of that earlier work was wasted: hard-won shared business definitions carried directly into the new dbt models, and old CTEs translated cleanly into staging and intermediate layers.
- 03
One outbound-communication script that took three and a half hours to run against a 14-billion-row table is now a layered set of dbt models an analyst can query selectively instead of scanning end to end.
Session speakers
Kensly Alexander
Planned Parenthood Federation of AmericaAriel Kaplan
Planned Parenthood