Administers the cloud data warehouse at the platform level: warehouse sizing and auto-suspension policy (Snowflake), slot and reservation management (BigQuery), partition and clustering strategies, query-cost attribution, role-based access control, and backup/restore procedures. Distinct from tech-dim.dev-team.backend-data-engineering, which builds the pipelines that run on the warehouse; this team ensures the platform itself is performant, cost-governed, and secure.
Why this is a separate discipline. In organizations without a dedicated warehouse administration function, compute costs spiral as unconstrained warehouses run idle, access control drifts and creates compliance exposure, and long-running composable-stack queries consume disproportionate resources without attribution. Composable CDP economics depend on this function: the cost advantage over packaged CDPs evaporates when warehouse spend is ungoverned.
When this team is absent. Organizations typically experience cost overruns within 6 to 12 months of going composable at scale. The absence of this function is a leading indicator that a composable architecture will exceed budget projections.