Skip to main content

Zendesk

Zendesk

Integrate your ept AI chatbot with Zendesk to provide intelligent responses to support tickets, access historical ticket data, and leverage your Zendesk knowledge base.

Back to Integrations Overview

Overview

The Zendesk integration enables your ept AI chatbot to:

  • Automatically respond to support tickets using your knowledge base
  • Access anonymous historical ticket data for context and learning
  • Leverage your Zendesk knowledge base for accurate responses
  • Provide 24/7 automated support with human escalation when needed

The integration uses secure OAuth authentication and follows the standard ept AI integration pattern: create an integration in the ept AI app, then configure a channel to use that integration.

Prerequisites

Before setting up the Zendesk integration, ensure you have:

  • Zendesk Account: Active Zendesk Support account with admin access
  • ept AI Setup: Your ept AI chatbot configured and ready
  • Knowledge Base: Support documentation loaded into ept AI

Setup Instructions

Step 1: Create Zendesk Integration in ept AI

  1. Access Integration Settings:

    • Log into your ept AI dashboard
    • Navigate to Configuration > Integrations
    • Click "Create Integration"
    • Select "Zendesk" from the available integrations
  2. Authorize Connection:

    • Click "Connect to Zendesk"
    • Complete the OAuth authorization flow
    • Grant necessary permissions to ept AI
    • The integration will be automatically configured

Step 3: Configure Zendesk Channel

  1. Create Channel:

    • Navigate to Configuration > Channels
    • Click "Create Channel"
    • Configure the channel:
      • Channel Name: "Zendesk Support"
      • Channel Type: Select "Zendesk"
      • Integration: Select the Zendesk integration you created
      • Knowledge Source Configuration: Choose support-focused KSC
  2. Configure Channel Settings:

    {
    "zendesk_channel_config": {
    "integration_id": "zendesk_integration_123",
    "knowledge_source": "support-ksc",
    "auto_respond": true,
    "escalation_threshold": 0.7,
    "access_historical_tickets": true,
    "use_zendesk_knowledge_base": true,
    "response_templates": {
    "greeting": "Thank you for contacting support. I'm here to help!",
    "escalation": "I'm transferring your ticket to a human agent who will assist you shortly."
    }
    }
    }

Integration Features

Ticket Response Automation

  • Automatic Responses: AI responds to new tickets using your knowledge base
  • Context Awareness: Uses ticket content and customer information for personalized responses
  • Escalation Management: Routes complex issues to human agents when needed
  • Response Quality: Maintains professional tone and accuracy

Historical Ticket Access

  • Anonymous Data: Access to historical ticket data for context and learning
  • Pattern Recognition: Identify common issues and improve response accuracy
  • Trend Analysis: Understand support patterns and optimize responses
  • No PII: Historical data is anonymized for privacy compliance

Zendesk Knowledge Base Integration

  • Article Access: Leverage your existing Zendesk knowledge base articles
  • Content Sync: Use Zendesk articles as part of your AI knowledge base
  • Consistent Information: Ensure responses align with your documented processes
  • Content Updates: Automatically incorporate knowledge base updates
  • Search and Access: Search help center articles and guides for intelligent responses
  • Public and Internal Articles: Access both public and internal knowledge base content
  • Category Filtering: Filter articles by categories for more relevant responses
  • Content Extraction: Process article content and attachments for comprehensive answers

Configuration Options

Ticket Response Settings

{
"ticket_response": {
"auto_respond": true,
"escalation_threshold": 0.7,
"response_templates": {
"greeting": "Thank you for contacting support. I'm here to help!",
"escalation": "I'm transferring your ticket to a human agent who will assist you shortly."
}
}
}

Knowledge Base Settings

{
"knowledge_base": {
"search_scope": {
"public_articles": true,
"internal_articles": true,
"draft_articles": false
},
"article_categories": ["how-to", "troubleshooting", "faq", "guides"],
"content_extraction": {
"enabled": true,
"include_attachments": true,
"max_article_length": 50000
},
"search_options": {
"fuzzy_matching": true,
"category_filtering": true,
"relevance_threshold": 0.7
}
}
}
{
"ticket_response_config": {
"auto_respond": true,
"response_mode": "suggested", // "auto", "suggested", "manual"
"escalation_triggers": [
"complex_issue",
"customer_request",
"low_confidence",
"urgent_priority"
],
"max_response_time": "5 minutes",
"include_knowledge_base": true
}
}

Historical Data Access

{
"historical_data_config": {
"enabled": true,
"data_retention": "90 days",
"anonymization": true,
"access_patterns": [
"ticket_categories",
"resolution_times",
"common_issues"
]
}
}

Knowledge Base Integration

{
"knowledge_base_config": {
"use_zendesk_articles": true,
"sync_frequency": "daily",
"article_categories": ["support", "faq", "troubleshooting"],
"content_priority": "zendesk_first" // "zendesk_first", "ept_first", "combined"
}
}

Use Cases

Automated Support

  1. Instant Responses: Provide immediate answers to common support questions
  2. Ticket Classification: Automatically categorize and prioritize tickets
  3. Escalation Management: Route complex issues to human agents
  4. Resolution Tracking: Track ticket resolution times and customer satisfaction

Knowledge Management

  • Content Optimization: Identify gaps in knowledge base coverage
  • Response Improvement: Learn from historical ticket patterns
  • Process Refinement: Optimize support workflows based on data
  • Training Enhancement: Use historical data to improve AI responses
  • Self-Service Support: Provide instant answers from knowledge base articles
  • Article Recommendations: Suggest relevant help center articles to customers
  • Content Updates: Keep AI responses current with latest knowledge base content
  • Agent Assistance: Help support agents find relevant articles quickly
  • Knowledge Gap Identification: Identify missing content in knowledge base
  • Content Optimization: Improve article discoverability and relevance

Customer Experience

  • Faster Resolution: Reduce response times with automated assistance
  • Consistent Information: Ensure all customers receive accurate, up-to-date information
  • 24/7 Support: Provide round-the-clock support availability
  • Personalized Responses: Tailor responses based on customer history and context

Best Practices

Integration Management

  1. Start Small: Begin with a limited set of ticket types
  2. Monitor Performance: Track response accuracy and customer satisfaction
  3. Gradual Expansion: Expand to more ticket categories based on success
  4. Regular Review: Periodically review and optimize integration settings

Data Privacy

  1. Anonymization: Ensure historical data is properly anonymized
  2. Access Control: Limit access to sensitive ticket information
  3. Compliance: Follow data protection regulations and company policies
  4. Audit Logging: Maintain logs of all data access and processing

Response Quality

  1. Knowledge Base: Keep Zendesk knowledge base articles updated
  2. Training Data: Use historical tickets to improve AI responses
  3. Escalation Rules: Set appropriate thresholds for human escalation
  4. Feedback Loop: Collect and incorporate customer feedback

Troubleshooting

Common Issues

OAuth Authentication Errors

  • Verify Client ID and Client Secret are correct
  • Check redirect URI matches exactly
  • Ensure OAuth app is properly configured in Zendesk
  • Verify required scopes are granted

Integration Connection Issues

  • Check integration configuration in ept AI
  • Verify OAuth connection is active
  • Test connection outside of channel configuration
  • Review error logs for specific issues

Channel Configuration Problems

  • Verify integration is properly linked to channel
  • Check knowledge source configuration
  • Review channel settings and permissions
  • Test with sample tickets first

Error Codes

Error CodeDescriptionSolution
401UnauthorizedCheck OAuth credentials and token
403ForbiddenVerify API permissions and scopes
404Not FoundCheck Zendesk object IDs and URLs
429Rate LimitedImplement rate limiting and retry logic

Support Resources

Monitoring & Analytics

Key Metrics

  • Ticket Response Rate: Percentage of tickets with AI responses
  • Escalation Rate: Percentage of tickets escalated to humans
  • Resolution Time: Average time to resolve AI-assisted tickets
  • Customer Satisfaction: CSAT scores for AI responses

Dashboard Views

Configure your ept AI dashboard to show:

  • Zendesk integration status
  • Recent ticket activity and responses
  • Historical data access patterns
  • Knowledge base usage statistics

Next Steps

After setting up your Zendesk integration:

  1. Test the Integration: Create test tickets and verify AI responses
  2. Monitor Performance: Track key metrics and optimize configurations
  3. Train Your Team: Ensure support agents understand the new workflow
  4. Scale Gradually: Expand to additional ticket types based on success

For additional support or custom Zendesk configurations, contact our integration team.