Automated Alerts

Alerts automatically notify you when something goes wrong with your AI agents—whether it’s a quality issue, compliance violation, or technical problem. Instead of manually monitoring every call, let alerts catch issues for you.

Why Alerts Matter

You can’t watch every conversation. Alerts help you:

Catch problems immediately - Get notified within seconds
Focus on what matters - Only review flagged calls, not every single one
Prevent escalations - Fix issues before customers get more frustrated
Ensure compliance - Know immediately when regulations aren’t followed

Example: An alert notifies you that Agent V1 has missed the required call recording disclosure in 3 calls today. You fix the prompt immediately before more violations occur.

How Alerts Work

You define conditions to monitor (e.g., “score drops below 70”), and Chanl watches all calls for those conditions. When triggered, you’re notified via your chosen channel.

Call happens → Monitored by alert rules → Condition met → Notification sent

Creating Alerts

Via UI
Via API

Navigate to Alerts and click “Create Alert”:

Name your alert
Define the condition to monitor
Set severity level
Choose notification channels
Activate the alert

curl -X POST https://api.chanl.ai/v1/alerts \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Low Quality Score Alert",
    "condition": {
      "type": "score_threshold",
      "threshold": 70,
      "operator": "less_than"
    },
    "severity": "high",
    "notifications": {
      "email": ["team@company.com"],
      "slack": {
        "webhook": "https://hooks.slack.com/...",
        "channel": "#ai-alerts"
      }
    },
    "agents": ["agent-v1", "agent-v2"],
    "active": true
  }'

Alert Types

Quality Alerts

Monitor agent performance quality:

Score Drop

Triggers when call quality score falls below threshold.

{
  "name": "Quality Drop Alert",
  "condition": {
    "type": "score_threshold",
    "threshold": 75,
    "operator": "less_than"
  },
  "severity": "medium"
}

When to use: Catch underperforming conversations immediately

Customer Frustration

Detects negative sentiment or frustration indicators.

{
  "name": "Customer Frustration Detected",
  "condition": {
    "type": "sentiment",
    "value": "negative",
    "confidence": 0.8
  },
  "severity": "high"
}

When to use: Identify calls that may need human intervention

Long Call Duration

Triggers when calls exceed expected length.

{
  "name": "Call Taking Too Long",
  "condition": {
    "type": "duration",
    "threshold": 600,
    "operator": "greater_than"
  },
  "severity": "medium"
}

When to use: Find agents stuck or unable to resolve issues

Compliance Alerts

Ensure regulatory requirements are met:

Missing Disclosure

Triggers when required disclosures aren’t provided.

{
  "name": "Missing Call Recording Disclosure",
  "condition": {
    "type": "required_phrase",
    "phrase": "This call may be recorded",
    "location": "first_30_seconds",
    "must_include": true
  },
  "severity": "critical"
}

When to use: Regulatory compliance monitoring

Policy Violation

Detects when agents violate company policies.

{
  "name": "Refund Policy Violation",
  "condition": {
    "type": "custom_rule",
    "prompt": "Agent offered refund outside 30-day policy window without manager approval"
  },
  "severity": "high"
}

When to use: Prevent unauthorized exceptions

Data Privacy Issue

Alerts when sensitive data is mishandled.

{
  "name": "PII Exposure Risk",
  "condition": {
    "type": "sensitive_data",
    "detect": ["ssn", "credit_card", "password"],
    "action": "mentioned_in_clear"
  },
  "severity": "critical"
}

When to use: Protect customer privacy

Technical Alerts

Monitor system health and errors:

Tool Failure

Triggers when agent tools fail to execute.

{
  "name": "CRM Tool Failure",
  "condition": {
    "type": "tool_error",
    "toolName": "lookup_customer",
    "consecutiveFailures": 3
  },
  "severity": "high"
}

When to use: Detect integration issues early

Timeout

Alerts when agent takes too long to respond.

{
  "name": "Agent Response Timeout",
  "condition": {
    "type": "response_time",
    "threshold": 10,
    "operator": "greater_than"
  },
  "severity": "high"
}

When to use: Identify performance bottlenecks

High Error Rate

Triggers when errors occur frequently.

{
  "name": "Elevated Error Rate",
  "condition": {
    "type": "error_rate",
    "threshold": 0.1,
    "timeWindow": "1h",
    "operator": "greater_than"
  },
  "severity": "critical"
}

When to use: Detect system-wide problems

Notification Channels

Email Notifications

{
  "notifications": {
    "email": {
      "recipients": [
        "team@company.com",
        "oncall@company.com"
      ],
      "subject": "Alert: {{alert_name}}",
      "includeCallLink": true,
      "includeTranscript": true
    }
  }
}

Slack Integration

{
  "notifications": {
    "slack": {
      "webhook": "https://hooks.slack.com/services/YOUR/WEBHOOK/URL",
      "channel": "#ai-agent-alerts",
      "mentionOnCritical": "@oncall",
      "includeActions": true
    }
  }
}

The Slack notification includes:

Alert details
Call information
Quick action buttons (View Call, Transfer, Dismiss)
Link to full analysis

Webhook Integration

{
  "notifications": {
    "webhook": {
      "url": "https://your-system.com/chanl-alerts",
      "method": "POST",
      "headers": {
        "Authorization": "Bearer YOUR_TOKEN",
        "X-Source": "chanl"
      },
      "payload": {
        "alertId": "{{alert_id}}",
        "callId": "{{call_id}}",
        "severity": "{{severity}}",
        "timestamp": "{{timestamp}}"
      }
    }
  }
}

SMS Notifications

{
  "notifications": {
    "sms": {
      "numbers": ["+1234567890"],
      "onlyForSeverity": ["critical"],
      "message": "Alert: {{alert_name}} - Call: {{call_id}}"
    }
  }
}

SMS notifications should be reserved for critical alerts only to avoid alert fatigue.

Alert Severity Levels

Low

Minor issues that don’t require immediate action. Review when convenient.Examples: Slightly low score, minor inconsistency

Medium

Issues that should be addressed soon. Review within hours.Examples: Score below 70, customer mild frustration

High

Problems requiring prompt attention. Review within minutes.Examples: Major compliance miss, angry customer

Critical

Urgent issues needing immediate action. Review immediately.Examples: Data breach risk, complete system failure

Managing Alert Noise

Throttling

Prevent alert spam by setting cooldown periods:

{
  "throttling": {
    "enabled": true,
    "cooldownMinutes": 15,
    "maxAlertsPerHour": 10
  }
}

This prevents the same alert from triggering more than once every 15 minutes.

Grouping

Combine similar alerts into batches:

{
  "grouping": {
    "enabled": true,
    "timeWindow": "5m",
    "maxGroupSize": 5,
    "notificationFormat": "{{count}} calls triggered '{{alert_name}}' in last 5 minutes"
  }
}

Instead of 5 separate notifications, you get one: “5 calls triggered ‘Low Quality Score’ in last 5 minutes.”

Scheduled Silence

Disable alerts during specific times:

{
  "silenceSchedule": {
    "enabled": true,
    "windows": [
      {
        "days": ["saturday", "sunday"],
        "startTime": "00:00",
        "endTime": "23:59"
      }
    ]
  }
}

Viewing Triggered Alerts

Via UI
Via API

Navigate to Alerts to see:

List of recent alerts
Filter by severity or agent
Click alert to view associated call
Mark as resolved or dismiss

# Get recent alerts
curl https://api.chanl.ai/v1/alerts/triggered \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "timeRange": "24h",
    "severity": ["high", "critical"]
  }'

Response:

{
  "alerts": [
    {
      "id": "alert_trigger_123",
      "alertId": "alert_abc",
      "alertName": "Low Quality Score",
      "callId": "call_xyz",
      "triggeredAt": "2024-01-15T10:30:00Z",
      "severity": "high",
      "resolved": false,
      "details": {
        "score": 68,
        "threshold": 70,
        "agent": "agent-v1"
      }
    }
  ],
  "total": 23
}

Alert Analytics

Track alert patterns over time:

const chanl = require('@chanl/sdk');

// Get alert frequency by type
const analytics = await chanl.alerts.analytics({
  timeRange: '30d',
  groupBy: 'alert_type'
});

console.log(analytics);
/*
{
  period: "30 days",
  totalAlerts: 147,
  byType: [
    {
      type: "quality_drop",
      count: 67,
      percentage: 45.6,
      trend: "decreasing"
    },
    {
      type: "compliance_issue",
      count: 42,
      percentage: 28.6,
      trend: "stable"
    },
    {
      type: "tool_failure",
      count: 38,
      percentage: 25.8,
      trend: "increasing"
    }
  ]
}
*/

Best Practices

Start with Critical Alerts

Begin with compliance and critical quality alerts. Add more as you understand patterns.

Set Appropriate Thresholds

Don’t make thresholds too sensitive. Test and adjust based on actual alert volume.

Use Throttling

Always enable throttling to prevent alert fatigue from repeated similar issues.

Review and Refine

Weekly review which alerts are useful vs noisy. Disable or adjust unhelpful ones.

Document Response Procedures

Create runbooks for each critical alert type so team knows exactly what to do.

Troubleshooting

Not receiving alerts

Problem: Alerts configured but no notifications arrivingSolutions:

Verify alert is marked as “active”
Check notification channels are configured correctly
Test with webhook/email validator
Confirm calls are actually triggering conditions
Check spam folder for email notifications

Too many alerts

Problem: Being overwhelmed with alert notificationsSolutions:

Enable throttling with appropriate cooldown
Increase severity thresholds
Use alert grouping to batch notifications
Review if conditions are too broad
Disable low-value alerts

Alerts triggering incorrectly

Problem: Getting alerts for false positivesSolutions:

Review condition logic - may be too broad
Adjust thresholds based on actual data
Add additional conditions to narrow scope
Test alert with historical calls first

Slack notifications failing

Problem: Slack alerts not appearingSolutions:

Verify webhook URL is still valid
Check Slack app permissions
Confirm channel exists and bot has access
Test webhook with curl to isolate issue

What’s Next?

Live Calls

Monitor calls that triggered alerts in real-time

Call Logs

Review historical alert patterns

Analytics

See alert trends and frequency

Improve Agents

Fix issues causing repeated alerts

Overview

Test

Observe

Optimize

Alerts

Automated Alerts

Why Alerts Matter

How Alerts Work

Creating Alerts

Alert Types

Quality Alerts

Compliance Alerts

Technical Alerts

Notification Channels

Email Notifications

Slack Integration

Webhook Integration

SMS Notifications

Alert Severity Levels

Low

Medium

High

Critical

Managing Alert Noise

Throttling

Grouping

Scheduled Silence

Viewing Triggered Alerts

Alert Analytics

Best Practices

Troubleshooting

What’s Next?

Live Calls

Call Logs

Analytics

Improve Agents

Overview

Test

Observe

Optimize

​Automated Alerts

​Why Alerts Matter

​How Alerts Work

​Creating Alerts

​Alert Types

​Quality Alerts

​Compliance Alerts

​Technical Alerts

​Notification Channels

​Email Notifications

​Slack Integration

​Webhook Integration

​SMS Notifications

​Alert Severity Levels

Low

Medium

High

Critical

​Managing Alert Noise

​Throttling

​Grouping

​Scheduled Silence

​Viewing Triggered Alerts

​Alert Analytics

​Best Practices

​Troubleshooting

​What’s Next?

Live Calls

Call Logs

Analytics

Improve Agents

Automated Alerts

Why Alerts Matter

How Alerts Work

Creating Alerts

Alert Types

Quality Alerts

Compliance Alerts

Technical Alerts

Notification Channels

Email Notifications

Slack Integration

Webhook Integration

SMS Notifications

Alert Severity Levels

Managing Alert Noise

Throttling

Grouping

Scheduled Silence

Viewing Triggered Alerts

Alert Analytics

Best Practices

Troubleshooting

What’s Next?