Implement dbt exposures to document and track downstream BI dashboards and data consumers dependent on dbt models

Steps

1. Create an exposures.yml file in your models directory; define an exposure with name, type (dashboard, notebook, analysis, ml, application, or other), owner (name and email), and depends_on listing the dbt models or sources it consumes: depends_on: - ref('fct_orders') - ref('dim_customers')
2. Add optional metadata fields: label (display name), description (what the exposure is for), url (link to the BI dashboard or notebook), maturity (high, medium, low) to indicate how production-ready the exposure is
3. After adding exposures, run dbt docs generate and dbt docs serve to view the lineage graph; exposures appear as terminal nodes in the DAG showing which dbt models feed which downstream consumers
4. Use exposures in CI to scope dbt builds: dbt build --select +exposure:my_dashboard builds all upstream models and sources that feed the specified exposure, ensuring nothing upstream breaks before the dashboard is affected
5. Keep exposures up to date as dashboards evolve; stale exposure definitions that reference deleted models will cause dbt parse errors — treat exposure files as living documentation with the same rigor as model YAML files