AI Performance Metrics Dashboard Architect

You are an Enterprise AI Governance Consultant specializing in performance optimization and business intelligence integration. Design a comprehensive AI Performance Metrics Dashboard for [AI_SYSTEM_TYPE] operating within [BUSINESS_CONTEXT].

STRUCTURE REQUIREMENTS:

1. EXECUTIVE COMMAND CENTER (Strategic View)
   - Define 4-6 high-level business impact metrics (e.g., AI-Driven Revenue Attribution, Cost Avoidance, Productivity Gains, Risk Mitigation Score)
   - Set Red/Yellow/Green threshold parameters with specific numerical targets
   - Include a 'Confidence Score' aggregating model reliability and data quality

2. OPERATIONAL HEALTH MONITOR (Technical View)
   - Performance metrics: Latency (p50, p95, p99), Throughput (requests/sec), Uptime/Availability
   - Model integrity: Prediction drift, Data freshness, Feature importance stability, Bias detection indices
   - Error analysis: Failure categorization (input errors, model errors, system errors), Retry rates

3. FINANCIAL PERFORMANCE PANEL
   - Cost per prediction/inference breakdown (compute, API, storage)
   - Total Cost of Ownership (TCO) vs. Legacy Process Comparison
   - Resource utilization efficiency (GPU/CPU usage patterns, idle time costs)

4. USER ADOPTION & EXPERIENCE LAYER
   - Adoption velocity: Daily/Monthly Active Users, Feature utilization rates
   - Human-AI collaboration metrics: Acceptance rate of AI recommendations, Override frequency, Time-to-decision improvements
   - Satisfaction indicators: NPS scores specific to AI interactions, Support ticket reduction rates

5. GOVERNANCE & COMPLIANCE TRACKER
   - Explainability audit log accessibility
   - Fairness/bias metrics across demographic segments
   - Data privacy compliance status (PII detection, consent management)
   - Model versioning and lineage documentation links

6. VISUALIZATION & INTERACTION DESIGN
   - Specify chart types for each metric category (e.g., 'Use heatmaps for bias detection across regions', 'Sparklines for 30-day trend micro-charts')
   - Define refresh frequencies (real-time vs. hourly vs. daily) per metric sensitivity
   - Design drill-down paths (e.g., 'Click ROI metric → See cost breakdown by department')

STAKEHOLDER CUSTOMIZATION:
Tailor the dashboard complexity for [STAKEHOLDER_AUDIENCE]:
- If Executive: Emphasize financial outcomes and risk posture with minimal technical jargon
- If Technical: Include infrastructure metrics, log correlation IDs, and debugging pathways
- If Operational: Focus on workflow integration points and exception handling queues

CONTEXT SPECIFICATIONS:
- Primary Business Objective: [BUSINESS_OBJECTIVE]
- Reporting Cadence: [REPORTING_FREQUENCY]  
- Current AI Maturity Stage: [MATURITY_LEVEL]
- Critical Success Factor: [KEY_SUCCESS_METRIC]

DELIVERABLE FORMAT:
Present as a dashboard specification document with:
- Wireframe descriptions for each panel (position, size, priority)
- Data source mapping (where each metric originates: API logs, database, user surveys)
- Alert configuration rules (thresholds triggering notifications to specific roles)
- Action playbook ('If Metric X drops below Y for Z hours, execute Procedure A')