Reverse ETL: Warehouse Data Back to Sales Tools

A SaaS company had built a solid data warehouse in Snowflake, but the insights were trapped there. The sales team used HubSpot, the customer success team used Intercom, and the marketing team used Braze. None of these tools had access to the rich customer data sitting in the warehouse. Sales reps were manually looking up customer health scores. Marketing was sending campaigns without segmentation data.

I built a reverse ETL pipeline that pushed curated data from Snowflake back into the operational tools where teams actually work. The system synced customer health scores, product usage metrics, churn risk flags, and segmentation tags directly into HubSpot, Intercom, and Braze.

Apache Airflow orchestrated the sync schedules. dbt models prepared the outbound datasets with the exact schema each destination tool expected. Python scripts handled the API integrations with rate limiting, error handling, and retry logic.

Every sync was idempotent and tracked. If a record failed to sync, it was logged and retried automatically. A monitoring dashboard showed sync health across all destinations in real-time.

Sales team got customer health scores directly in HubSpot, eliminating manual lookups

Marketing campaigns became data-driven with warehouse-powered segmentation

Customer success team saw churn risk flags in Intercom before renewal conversations

Time-to-insight for operational teams dropped from days to minutes

Zero data sync failures in 3 months of production operation

Snowflake, dbt, Apache Airflow, Python, SQL

Solo data engineer. Designed the reverse ETL architecture, built all API integrations, and deployed the monitoring system. Delivered in 6 weeks.