Real-Time Business Intelligence Agent

# Real-Time Business Intelligence Agent

You are a data engineering and business intelligence architect. Design an agent that monitors [BUSINESS_TYPE] metrics in real time.

## Business Context

- Business type: [BUSINESS_TYPE]
- Key metrics to monitor: [METRIC_LIST]
- Current monitoring gaps: [CURRENT_GAPS]
- Stakeholders who need alerts: [STAKEHOLDER_LIST]

## Metric Definitions

For each metric, define:
- Metric name and data source
- Update frequency (real-time / 5-min / hourly / daily)
- Healthy range, warning threshold, critical threshold
- Business impact if it breaches critical

## Anomaly Detection Rules

**Statistical Anomalies**
- Spike: metric exceeds [SPIKE_MULTIPLIER]x the rolling [WINDOW] average
- Drop: metric falls below [DROP_THRESHOLD]% of baseline
- Trend break: 3 consecutive readings outside the normal band

**Business Logic Anomalies**
- [BUSINESS_RULE_1]
- [BUSINESS_RULE_2]
- [BUSINESS_RULE_3]

## Alert Routing

| Severity | Condition | Recipients | Response Time |
|----------|-----------|------------|---------------|
| P1 | [CONDITION] | [RECIPIENTS] | 15 minutes |
| P2 | [CONDITION] | [RECIPIENTS] | 2 hours |
| P3 | [CONDITION] | [RECIPIENTS] | Next business day |

## Alert Content Template

Every alert includes:
- Metric name and current value
- Expected range and threshold breached
- Time the anomaly was first detected
- Top 2 likely causes
- Recommended immediate action
- Dashboard link

## Root Cause Analysis Sub-Agent

Design a sub-agent that activates on P1 alerts and surfaces the most likely root cause within [INVESTIGATION_WINDOW] minutes.