Common AI Data Explorer issues and fixes

  • Release version: Yokohama
  • Updated March 25, 2026
  • 1 minute to read

    • Use Query Generation logs to inspect failed or incorrect queries and identify common issues with AI Data Explorer responses.

    Go to Query Generation > Administration > Logs to inspect failed or incorrect queries.
    1. Find the entry by user, date, or utterance.
    2. Check Semantic Entities/Dimensions/Segments Count
    3. Open Query Constitutor Output to see what was selected.
    4. Compare with Semantic Filter Result for raw semantic matches.
    Table 1. AI Data Explorer troubleshooting guide
    Symptom Likely cause Fix
    "Unable to understand" response System cannot find a matching table Improve entity description; rephrase with table name; verify AI search is running; verify that all Indexed Sources are fully processed.
    Wrong table selected Similar table names, no disambiguation Add semantic usage instructions to differentiate.
    Wrong table selected (permissions) User lacks read access to the intended table or fields Verify and grant table and field read ACLs for the intended data.
    Missing field in results Dimension turned off, unclear label, or field not in semantic layer Check dimension is active; update label.
    Related fields not pulled in Referenced table missing from semantic layer or reference dimension turned off Add and enable the referenced table entity; enable the reference dimension.
    Wrong filter applied Bad segment match Fix segment name and description; disable the segment.
    No filter when expected No matching segment; phrasing too vague Create a manual segment; rephrase question.
    Segment from report is wrong Auto-generated segment has stale or irrelevant filter Disable the report segment source or the specific segment.