Partnership: Snowflake
Warehouse partnership.
Overview
Snowflake powers many customer analytics stacks. Operational data (metrics, logs, traces) and analytics data (product, business) increasingly overlap, and the integration produces value at the seam: ops queries against the warehouse, Snowpipe streaming for near-real-time, role-based auth that preserves the existing security model.
- Snowflake as a source. Read operational metrics, logs, traces from Snowflake tables. Warehouse-first patterns work.
- Snowpipe ingestion. Stream operational events into Snowflake. Near-real-time analytics on operational data.
- Standard SQL. Snowflake SQL drives Nova queries. Matches data-team workflows.
- Role-based auth plus cost-aware queries. Snowflake RBAC preserves the security model; bounded query patterns avoid runaway warehouse cost.
The approach
Three habits make warehouse integration produce real value: Snowflake as source for analytical workloads, Snowpipe for streaming where freshness matters, role-based authentication so the existing security model carries over.
- Snowflake as source. Operational data lands in Snowflake. Warehouse-first patterns work.
- Snowpipe for streaming. Near-real-time ingestion supports fast detection.
- Role-based authentication. Snowflake RBAC controls access. Security model preserved.
- Cost-aware queries plus documented integration. Bounded query patterns avoid warehouse cost surprises; per-integration the setup guide for self-service.
Why this compounds
Each integration deposits cross-team collaboration. Operations and data teams share substrate; historical analysis becomes queryable across years; existing Snowflake investment gets leveraged rather than duplicated.
- Cross-team collaboration. Ops and data share data. Real collaboration follows.
- Historical analysis. Long-term operational data queryable. Trend analysis becomes possible.
- Preserved investment. Existing Snowflake setup leveraged. Adoption stays low-friction.
- Year-one investment, year-two habit. First integration is investment. By the third customer, the integration template is settled.