Stream Golden DB into your Snowflake, BigQuery, or Databricks warehouse so downstream analytics, ML models, and BI tools work from the same canonical data the agents use.
Why warehouse export matters
Your data warehouse becomes the analytical source of truth when it has Golden DB. Without the export, your BI tools work from CRM data (partial and stale) or from individual provider data (fragmented). The export gives you one queryable source for accounts, contacts, signals, and agent activity.
What gets exported
Account, Contact, Deal, Signal, and Activity tables with the full Golden Record schema. Provenance is summarized (full provenance lives in turgo for performance reasons). Standard configuration exports incremental updates every 15 minutes; near-real-time configuration runs streaming exports under 1 minute latency.
Setting up Snowflake export
Settings → Data → Warehouse Export → Snowflake → enter account, warehouse, database, schema → authorize via OAuth or key pair. turgo creates the target tables on first run; subsequent runs incremental-merge new and updated records.
Setting up BigQuery export
Same flow, BigQuery panel: project ID, dataset name, service account JSON. The first export creates partitioned tables; subsequent exports append partitions.
Setting up Databricks export
Same flow, Databricks panel: workspace URL, catalog name, schema name, personal access token or service principal. Delta Lake tables are created; incremental updates merge via Delta MERGE.
Validating the export
Run a simple count query in your warehouse and compare to turgo's Account, Contact, and Deal counts (visible in the dashboard). They should match within ±1 record due to in-flight events at sample time. If counts diverge significantly, check the export log for failed rows.
