Rollup Views

Pre-aggregated views built by the orchestrator for fast analytics queries — without scanning billions of rows in the full production table.

Available Rollups

View	Partition	Key Metrics	Typical Use
prod_rollup_provider	provider_id	Total ROIDs, rated ROIDs, avg rate, payer count	Provider rate landscape
prod_rollup_payer	payer_id	Total ROIDs, rated ROIDs, median rate, provider count	Payer rate behavior
prod_rollup_network	network_id	Median rate, provider count	Network-level competitiveness
prod_rollup_code	billing_code, bill_type	Total ROIDs, rated ROIDs, avg rate, provider count	National range for a service
prod_rollup_provider_network	provider_id, network_id	Per-contract rate statistics	Provider rate variation across networks
prod_rollup_cbsa_code	CBSA, billing_code	Regional pricing aggregates	Geographic pricing variation

Build Process

Rollups are built from prod_combined_abridged after the orchestrator merge completes. Each rollup applies uniqueness and non-null validation checks before writing — ensuring the aggregates reflect only valid, non-duplicate ROIDs.

When to use rollups vs the full table

Rollups are orders of magnitude faster for aggregate queries. Use prod_combined_abridged when you need rate-level detail for individual ROIDs. Use prod_combined_all when you need the full rate array or internal accuracy scores for debugging. Use rollup views when you need aggregates — national averages, provider coverage counts, network comparisons — without row-level detail.