```json { "slots": [ { "scope": "dataset", "slot": "nature", "value": { "nature": "operational" }, "evidence": { "reasoning": "This dataset tracks AI vs human call handling metrics from a running call center operation. The data measures outcomes of a business process (call routing and resolution) that the operator controls, including AI resolution rates, escalation patterns, and call volumes by day." } }, { "scope": "dataset", "slot": "decisions_served", "value": { "decisions": [ "Whether to adjust AI call routing thresholds to optimize resolution rates", "When to schedule additional human agents based on escalation patterns", "Whether the AI system is performing adequately or needs retraining", "How to allocate resources between AI and human handling capacity" ] }, "evidence": { "reasoning": "The dataset provides AI resolution percentages, human escalation rates, and call volumes that directly inform staffing and AI tuning decisions. The day-of-week and time patterns enable scheduling optimization." } }, { "scope": "dataset", "slot": "levers", "value": { "levers": [ "AI routing rules and escalation thresholds", "Human agent staffing levels and schedules", "AI model training and configuration", "Call handling protocols and scripts" ] }, "evidence": { "reasoning": "The operator controls the AI system configuration, staffing decisions, and call handling processes. The metrics track the outcomes of pulling these levers - AI resolution rates respond to model changes, human handling percentages respond to routing rules." } }, { "scope": "dataset", "slot": "audience", "value": { "audience": [ "Call center operations managers", "Workforce planning analysts", "AI/ML engineers responsible for call routing systems", "Customer service leadership" ] }, "evidence": { "reasoning": "Operations managers use this to monitor daily performance, workforce planners use it for staffing decisions, ML engineers use resolution rates to tune the AI, and leadership uses it to assess overall service delivery." } }, { "scope": "table:callyeah_analytics_2026_07_23", "slot": "role", "value": { "role": "fact" }, "evidence": { "reasoning": "This table records call handling events and their outcomes. Each row represents either a daily summary or individual call metrics, capturing measurable facts about call volume, resolution method, and performance." } }, { "scope": "table:callyeah_analytics_2026_07_23", "slot": "grain", "value": { "unit": "one row per call event or daily summary period", "key_columns": ["Day", "_ingestion_timestamp"] }, "evidence": { "reasoning": "The data appears to mix daily aggregates (rows with day-of-week like 'Thu', 'Fri') and potentially individual call records (rows with dates like '01-Jul', '02-Jul'). The 216 rows with high uniqueness in Day column (207 unique values) suggests a mix of granularities, with _ingestion_timestamp providing temporal ordering." } }, { "scope": "table:callyeah_analytics_2026_07_23", "slot": "temporal_posture", "value": { "posture": "typed", "columns": ["_ingestion_timestamp"] } }, { "scope": "table:callyeah_analytics_2026_07_23", "slot": "column_bindings", "value": { "bindings": [ {"column": "Day", "binding": "temporal"}, {"column": "Total Calls", "binding": "measure"}, {"column": "AI Handled", "binding": "measure"}, {"column": "Human Handled", "binding": "measure"}, {"column": "Voicemail", "binding": "identifier"}, {"column": "AI Resolution %", "binding": "measure"}, {"column": "Human %", "binding": "measure"}, {"column": "_unnamed_7", "binding": "measure"}, {"column": "_unnamed_8", "binding": "measure"}, {"column": "_unnamed_9", "binding": "category"}, {"column": "_ingestion_timestamp", "binding": "temporal"}, {"column": "_batch_id", "binding": "identifier"}, {"column": "_source_file", "binding": "identifier"}, {"column": "_source_system", "binding": "identifier"} ] }, "evidence": { "reasoning": "Day contains temporal values (day-of-week, dates). Total Calls, AI Handled, Human Handled are count measures. AI Resolution % and Human % are percentage measures. Voicemail appears to contain phone numbers (identifier/contact info). _unnamed_7 looks like duration (0:25, 0:33), _unnamed_8 like decimal measures. _unnamed_9 contains sentiment categories (Positive, Negative, Neutral). The derived binding for Total Calls, AI Resolution %, and Voicemail as 'category' is incorrect - these are measures or identifiers." } }, { "scope": "table:callyeah_analytics_2026_07_23", "slot": "derived_measures", "value": { "measures": [ { "column": "AI Resolution %", "derived_from": ["AI Handled", "Total Calls"], "rule": "AI Handled / Total Calls" }, { "column": "Human %", "derived_from": ["Human Handled", "Total Calls"], "rule": "Human Handled / Total Calls" } ] }, "evidence": { "reasoning": "AI Resolution % and Human % are clearly derived percentages calculated from the count columns. The sample values show these as percentages that would result from dividing handled counts by total calls." } } ] } ```