๐Ÿ“‹ Fabric Architecture Accelerator
Assessment Report
Client...
March 29, 2026 11:18 AM
gaston@thepowermates.com
Overall Health Score
74
Grade: C
1
CRITICAL
2
HIGH
2
MEDIUM
4
workspaces_designed
2
lakehouses_created
6
pipelines_scaffolded
8
notebooks_generated
Medallion (Bronze/Silver/Gold)
architecture_pattern

Findings (5)

No workspace isolation between dev/test/prod
CRITICAL
Architecture

All development, testing, and production artifacts exist in a single workspace. A bad deployment could break production dashboards.

โ†’ Implement 3-workspace pattern: Sales-Dev, Sales-Test, Sales-Prod with deployment pipelines between them.
Lakehouse schema not enforced
HIGH
Data Quality

Bronze layer tables lack schema enforcement, allowing malformed data to propagate to Silver and Gold layers.

โ†’ Add Delta table constraints and check constraints on Bronze ingestion. Use schema evolution with mergeSchema option.
Git integration not configured
HIGH
DevOps

No source control for notebooks, pipelines, or semantic models. Changes cannot be tracked, reviewed, or rolled back.

โ†’ Connect workspace to Azure DevOps or GitHub. Enable Git integration for all Fabric item types.
No data lineage documentation
MEDIUM
Governance

No documentation of data flow from source systems through Bronze โ†’ Silver โ†’ Gold layers.

โ†’ Generate lineage diagrams using Purview or manual documentation. Tag each table with source system and refresh frequency.
Missing capacity auto-scaling rules
MEDIUM
Cost

Fabric capacity is fixed at F64 with no auto-pause or scale rules. Evening/weekend usage is near zero but capacity remains allocated.

โ†’ Configure auto-pause after 15min idle. Set up scale rules: F64 during business hours, F16 evenings/weekends.

Recommendations

1Implement dev/test/prod workspace separation with deployment pipelines
2Enable Git integration for all workspaces
3Add schema enforcement on Bronze layer tables
4Configure capacity auto-scaling to reduce weekend/evening costs
5Document data lineage from source to Gold layer