Warehouse connections stream turgo data out for analytics (standard export) and can pull warehouse-computed attributes back in (reverse flow). Setup is credentials + destination + schedule.
Export setup
Settings → Data → Warehouse Export → pick platform. Snowflake: account/warehouse/database/schema + key pair or OAuth. BigQuery: project/dataset + service-account JSON. Databricks: workspace/catalog/schema + token. Tables auto-create on first run.
What lands
Accounts, contacts, deals, signals, and activity with full schema; incremental merge every 15 minutes (streaming option for near-real-time). Provenance summarized.
Reverse flow
Pull warehouse-computed columns (churn scores, product-usage tiers, LTV models) back onto turgo records: define the source view and key mapping, and the attributes become targetable in ICPs and workflows.
Validation
Compare row counts against the dashboard after the first sync (±1 tolerance for in-flight events). The export log itemizes any rejected rows.
