Score Dimensions
The AI Readiness score is built from thirteen scored dimensions. An assessment result shows one card per dimension. Each card states its count, such as 12 of 40 described, and its label opens an explanation of what the dimension measures and why it matters to an agent. The same explanations are collected here.
| Dimension | What it measures, and why it matters to an agent | Who fixes it |
|---|---|---|
| Description coverage | How many visible tables, columns, and measures carry a description. Where usage data is available, the objects your users engage with most count for the most. Where an AI data schema is set, only the included tables and fields are counted. | BI Pixie, through Table descriptions and Field descriptions |
| Name disambiguation | Measures or columns with confusingly similar names. Agents choose between fields by name, so near-duplicates produce wrong answers. A field that already carries a description is counted as clear, because BI Pixie never renames fields and a description is how a near-duplicate is told apart. Measures and columns are compared separately, since a measure often shares its name with the column it wraps. | BI Pixie, through Field descriptions |
| Field clarity | Names that are very short, all upper case, or underscored, which both agents and people struggle to interpret. Adding a description clears the finding. | BI Pixie, through Field descriptions |
| AI data schema | Whether the semantic model defines a focused subset of its schema for AI. A curated subset guides Copilot and data agents away from ambiguity and toward clearer, more accurate responses. | BI Pixie, through Curate the AI data schema |
| AI Instructions | Whether the semantic model carries AI Instructions from Prep data for AI. Instructions give agents business context the schema alone cannot convey: terminology, metric preferences, and how to route questions. | BI Pixie, through AI Instructions |
| Indexing limits | Whether the semantic model exceeds the number of fields Copilot in Power BI indexes. Past that limit, Copilot silently sees only part of the semantic model. Trimming the AI data schema keeps the relevant part visible. | BI Pixie, through Curate the AI data schema |
| Verified answers | Whether the semantic model carries Verified answers from Prep data for AI, and whether they cover the visuals your users engage with most. Verified answers pair common questions with approved visuals and are the most reliable way to make Copilot and data agents answer consistently. They can be authored only in Power BI Desktop, and Microsoft exposes no API that creates or edits one, so BI Pixie cannot automate this dimension as it does the others. That is also why it is weighted lightly: the findings show which visuals to start with, and the next assessment picks up whatever you author. | You, in Power BI Desktop |
| Model structure | Whether every table is related to the rest of the semantic model. Agents answer cross-table questions by following relationships, so an unrelated table is invisible to the queries they write. Relate orphan tables, or remove obsolete ones. | You |
| Time window safety | Measures that hard-code a fixed date range. An agent may present those results as though they covered all time. | You |
| PII and RLS safety | Columns that appear to contain personal data in semantic models that have no row-level security, including hidden columns and hidden tables. Hiding is not a security boundary; row-level or object-level security is what removes access. | You |
| Field defaults | Whether text columns summarize by default. A summarizing text default produces meaningless totals that confuse agents. | BI Pixie, through Fix field defaults |
| Display folders | Whether measures are organized into display folders, so that agents and report authors can navigate the semantic model. | BI Pixie, through Organize measures into folders |
| Implicit measures | Whether tables whose numeric columns auto-summarize also carry explicit measures. Without one, an agent must decide for itself whether to sum, average, or count the column, so the same question can return different numbers on different asks. Writing measures is modeling work, so these findings require manual changes in your semantic model. | You |
Reading a Dimension Card
Most cards show a count whose fraction matches the dimension's points. Four deliberately do not, and each says so when opened:
- Description coverage counts described fields evenly, while its points weight the heavily used fields more, so a semantic model can describe half its fields and score well above half if they are the right half.
- Indexing limits shows a budget rather than a tally: the number of fields Copilot indexes against the number the semantic model exposes, where a higher count is worse.
- Verified answers and PII and RLS safety show no count, because a presence check and a safety check are not fractions of anything.
Where usage data is available, the findings under a dimension carry badges such as Heavily used, On reports, Never clicked, and Not on any report, which is how BI Pixie ranks what to fix first in a grounded result. See Grounded in Real Usage.
What's Next
- Optimizations, the fixes BI Pixie writes for the dimensions it can change.
- Assessments, to run the assessment that produces these cards.