Table of Contents

AI DAX Debugger

The DAX Debugger uses AI to help you figure out why a measure is returning unexpected results. Instead of manually tracing filter context and referenced measures, you can ask the AI to analyze it and explain what's going on in plain English.

How to Use It

  1. Right-click a measure in the Explorer or Properties panel.
  2. Select Debug with AI.
  3. The AI analyzes the measure's DAX expression, its referenced measures, and the surrounding model structure (relationships, filter context paths).
  4. You get back:
    • An explanation of what the measure is doing and where the logic might be off
    • How filter context flows through the expression
    • Specific issues like unintended CALCULATE modifiers, incorrect filter propagation, or missing ALL wrappers
    • Suggested fixes with corrected DAX

What the AI Sees

The debugger sends only what's needed to understand the measure:

  • The measure's DAX expression
  • DAX expressions of any measures it references
  • Table and relationship metadata (names, types, cardinalities)

No raw data from your tables is ever sent.

AI Provider

The DAX Debugger uses whichever AI provider you've configured:

  • BYOK: Your own API key from Anthropic, OpenAI, Azure AI Foundry, Ollama, LM Studio, or any OpenAI-compatible endpoint. Included with every license.
  • AI Add-on: Hosted AI through Fact Foundry's proxy. No setup required. Available with the AI Add-on license tier.

Configure your provider in Settings > AI.

Edit this page