Tenant-wide settings, capacity, governance, audit, and usage for everyone in your organization.

Scale by SKU

Scale every backing service from inside Loom. Each card hits a real Azure REST endpoint via the BFF — no portal hand-offs. Cost previews are list-price estimates from a lookup table; reserved instances and SLA surcharges are excluded. Apply is async on most services (Fabric capacity, APIM, ADX) — refresh the card after a few minutes to see the new state.
Each card scales a real Azure backing service — no Microsoft Fabric capacity required. Scaling models differ by service:
  • Fabric / Power BI capacity — pick an F-SKU (F2 → F2048) or Power BI P-SKU; capacity units scale linearly and you can pause to stop compute billing.
  • Synapse dedicated SQL pool — a provisioned MPP warehouse scaled by DWU (DW100c → DW30000c). Cost scales with the DWU level; pause the pool when idle to pay for storage only. This is distinct from Synapse serverless, which has no dial — it autoscales and bills per TB scanned.
  • Cosmos DB — set throughput per container as manual RU/s (a fixed hourly reservation) or autoscale max RU/s (scales between 10% and 100% of the max on demand). Autoscale bills about 1.5× the manual rate per RU, but only for the RU/s actually used each hour — cheaper for spiky load, pricier for steady high load.
Fabric / Power BI CapacityF-SKU (F2 → F2048) for Fabric; P-SKU for Power BI Premium.
Utilization
Synapse Dedicated SQL Pool (DWU)DW100c → DW30000c — scale-out via ARM PATCH on sqlPools/{n}.
Utilization
Azure Data Explorer (ADX)vCore tier (Dev / E2 / E4 / E8 / E16 / E64) + capacity.
Utilization
Databricks SQL Warehousecluster_size (2X-Small → 4X-Large) via /api/2.0/sql/warehouses/{id}/edit.
Utilization
Databricks Clusternode_type_id + num_workers via /api/2.0/clusters/edit.
Utilization
Azure AI SearchSKU (S0/S1/S2/S3/S3HD) + replicas + partitions.
Utilization
API ManagementSKU (Developer / Basic / Standard / Premium / *V2) + capacity.
Utilization
Cosmos DB containersPer-container RU/s (manual) or autoscale max RU/s.
Utilization
Container Apps (Loom services)workload profile (Consumption / D-/E-series) + replicas.
Utilization
AI Foundry — AML computevmSize + min/max nodes for AmlCompute targets. Requires a Microsoft.MachineLearningServices/workspaces resource.
Utilization