Prompt Almanac

Audit advanced Works with: Any AI

Power BI Data Model Review for Performance and Correctness

Table output Has example

Check relationships, column choices and cardinality.

Open in ChatGPTOpen in Claude

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

[VERIFY] is an editorial flag, not a blank to fill. Wherever the prompt shows [VERIFY], check that item against a reliable source before you use it.

Rate this prompt

Loading ratings…

Review my Power BI data model. Tables with row counts and column counts: [TABLES]. Relationships (From, To, Cardinality, Direction): [RELATIONSHIPS]. Largest columns and their types: [LARGE_COLUMNS]. Symptoms (slow visuals, wrong totals, blank values): [SYMPTOMS].

Check for many-to-many relationships, bidirectional filters, high-cardinality text columns in fact tables, unused columns, calculated columns that could be measures, a missing date table, inactive relationships, and hidden duplicates in dimension keys. For each issue explain in plain words the effect on speed or accuracy, how to test it, and the fix, ranked by likely effect. Example (invented): 9 tables and 10 relationships, of which 2 are many-to-many and 1 is bidirectional.

Return a table with Issue, Evidence, Effect, Test and Fix, and a plan in an order that avoids breaking reports. Save a copy first, check totals before and after each change, and mark tool names such as performance analyser or model view [VERIFY]. Do not recommend changes needing data I did not provide.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

See an example result (fictional data)
Invented model. Tables: 9. Relationships: 10, of which 2 are many-to-many and 1 is bidirectional. Largest column: a text description of 3 million rows with 1.8 million distinct values (not used in any visual). Unused columns: 46 of 112 (41 percent). Date table: missing; 3 date columns each used in different visuals.
Fixes in order: remove unused columns (expected size reduction, measured), add a date table, replace the many-to-many relationships with a bridge or a corrected key, review the bidirectional relationship. Check: total sales stays 9,481,220.55 after each change.

Illustrative only, made-up data. Your results will differ and must be reviewed.

Related prompts