← All platforms

AI readiness · Microsoft Fabric

Microsoft Fabric AI readiness assessment

Score your Microsoft Fabric estate against the six factors that decide whether data is ready for an AI workload — read-only, metadata only, never a row of your data. Free to run.

Fabric estates are the youngest of any platform here, which cuts both ways. There is less accumulated residue than in a decade-old warehouse, and correspondingly less established convention about who documents what. Readiness on Fabric is usually less about cleaning up history and more about whether description is happening at all as the estate is built.

Both Warehouse and Lakehouse are assessed the same way, through the SQL analytics endpoint, using an Entra ID service principal with read-only workspace access.

The Microsoft Fabric readiness tension: newness — estates are young, conventions are still forming, and metadata sync is a real step.

What the assessment reads on Microsoft Fabric

• SQL analytics endpoint catalog — tables, columns, and types for both Warehouse and Lakehouse items.
• Declared constraints — keys and relationships, which determine entity resolution for a workload.
• Documentation coverage — descriptions at table and column level.
• Row-level security definitions — where VIEW DEFINITION has been granted — otherwise reported as not assessable.

Those signals roll up into the six factors — Clean, Contextual, Consumable, Current, Correlated, Compliant — scored against roughly 50 requirements for the specific AI workload you pick, whether that is retrieval, agents, model training, or feature serving. The output is a readiness score, the single biggest blocker, and a prioritised fix plan with the score gain each fix earns. See the methodology for how scoring works.

Access it needs

An Entra ID service principal added to the Fabric workspace with the Viewer role, plus the tenant setting "Service principals can use Fabric APIs" enabled. Viewer is read-only. Optionally, GRANT VIEW DEFINITION in the warehouse SQL editor unlocks the row-level-security signal; without it that one check reports not assessable rather than guessing.

The exact SQL each connector runs is published and generated from source, so a security review is a verification rather than a trust fall. If your warehouse cannot accept connections from the internet, the CLI agent runs the same metadata-only assessment inside your network. Full setup steps live in the connection guide.

The honest limits on Microsoft Fabric

• For a Lakehouse, newly created Delta tables appear in the SQL analytics endpoint only after Fabric's metadata sync — usually a few minutes. A scan run immediately after table creation can understate the estate.
• Without GRANT VIEW DEFINITION, the row-level-security requirement is not assessable. Everything else still scores.
• The tenant setting for service principals can take up to 15 minutes to propagate, which accounts for most first-attempt connection failures.

A metadata scan cannot verify values anywhere: a comment can be stale, a constraint can be worked around in application code, and a fresh-looking table can be full of duplicates. What it gives you is the structural truth of the estate — what exists, what is documented, what is governed, what is fresh, and what your chosen workload requires that is missing. It prepares evidence. It does not certify compliance, and no tool that reads only metadata honestly can.

Also covers: Fabric Warehouse, Fabric Lakehouse.

Frequently asked questions

Want to see the output before connecting anything? Read a sample readiness report built entirely from metadata. Other platforms: see all six.