dbt for public health: sightline wastewater monitoring at scale
Verily Health's five-person data team uses one reusable dbt package to monitor wastewater for more than 30 pathogens in near real time.
Overview
Verily's Sightline program tracks pathogens in wastewater across hundreds of US locations for early outbreak detection. Covers transforming nested FHIR data via composable dbt macros into dimensional models powering real-time dashboards, and scaling from one data mart to multiple healthcare implementations.
What you'll take away
- 01
A five-person team powers near real-time wastewater surveillance for 30-plus pathogens by building one shared dbt package with parameterized macros for each FHIR healthcare resource type, instead of solving the same modeling problem for two dozen separate datasets.
- 02
The same observation macro that tracks influenza A levels in a wastewater sample also tracks a person's A1C reading from a glucose monitor, with sensitivity and grain adjusted through macro parameters rather than duplicated code.
- 03
Two simple heuristics keep the resulting models usable: composite keys capped at around three columns, and most tables held to roughly 30 columns or fewer.
Session speaker
David Drummond
Verily Health