I'll systematically check for data quality issues across the dataset. Let me start by examining the column metadata to understand what we're working with. ```json { "action": "tool_call", "reasoning": "First, I need to see the detailed metadata for all columns in the main table to understand types, null counts, and value ranges. This will help identify potential outliers, impossible values, and data quality issues.", "tool": "inspect_columns", "args": { "table": "ai_job_dataset" } } ```