Architecture tool

Workload Placement Simulator

Decide where a workload should run before tuning compute, table layout, semantic models, access, or recovery.

Directional placement review

Use this when Databricks, Fabric, and Power BI are all plausible homes for part of the same workload. The output is a review brief, not a vendor sizing guarantee.

Do not paste confidential platform, customer, credential, or incident details.

Scenario

Describe the workload shape.

Pick the strongest signal for each dimension. The simulator ranks placement paths by fit and flags review risks before a team commits to a runtime.

Workload intent Batch or incremental processing with owned release and retry behavior. Platform estate Unity Catalog, Jobs, SQL Warehouse, notebooks, or Lakeflow are primary. Data shape Curated lakehouse tables are the main serving or processing surface. Freshness Freshness is measured in business cycles, not user interaction. Scale Small group, low concurrency, and limited operational blast radius. Governance Normal internal controls and moderate audit expectations. Dominant pressure Reduce cluster ownership, idle time, patching, and platform toil.

Workload Placement Simulator

Describe the workload shape.

Why this surfaced

Watchouts

Copyable summary