I need to systematically check for data quality issues across the dataset. Let me examine different aspects beyond the salary analysis I've already done. {"action": "tool_call", "reasoning": "Check temporal consistency - whether posting_date is always before application_deadline, and look for any impossible date relationships", "tool": "query_sql", "source": "raw", "purpose": "Identify temporal inconsistencies where application deadline precedes posting date", "args": {"sql": "SELECT COUNT(*) as invalid_date_count, MIN(posting_date) as earliest_posting, MAX(posting_date) as latest_posting, MIN(application_deadline) as earliest_deadline, MAX(application_deadline) as latest_deadline, MIN(DATEDIFF('day', posting_date, application_deadline)) as min_days_to_apply, MAX(DATEDIFF('day', posting_date, application_deadline)) as max_days_to_apply, AVG(DATEDIFF('day', posting_date, application_deadline)) as avg_days_to_apply FROM raw.ai_job_dataset WHERE application_deadline < posting_date"}}