Jira Escalations

​

Overview

​

Requirements

• Active Jira Cloud or Server instance
• Jira integration connected (see Integration Setup below)
• API token with issue creation permissions
• Jira project with appropriate issue types configured

​

Setup Instructions

​

Step 1: Connect Jira Integration

​

Step 2: Create Escalation Path

1. Navigate to Agents → Your Agent → Escalations
2. Click “Create Escalation Path”
3. Fill in the basic information:
   • Name: Descriptive name (e.g., “Jira Bug Reports”)
   • Platform: Select “Jira”
   • Platform Integration: Choose your connected Jira instance

​

Step 3: Configure Jira Settings

• Project Key: Jira project where issues will be created (e.g., “SUP”)
• Issue Type: Bug, Task, Story, Epic, etc.
• Priority (optional): Set default priority (Highest, High, Medium, Low, Lowest)
• Assignee (optional): Auto-assign to specific user
• Labels (optional): Add labels to issues (e.g., “ai-escalation”, “customer-reported”)
• Components (optional): Associate with project components

​

Step 4: Set Up Triggers (Optional)

• Manual Only: Only when customer requests or AI determines need
• Message Count: After specific number of messages
• Keywords: When certain words detected

​

Step 5: Save Configuration

​

Configuration Options

​

Project Key

• SUP for “Support”
• BUG for “Bug Tracking”
• CS for “Customer Success”

​

Issue Types

• Bug: Software defects reported by customers
• Task: Work items or action items
• Story: User stories or feature requests
• Epic: Large initiatives or themes
• Support: Support-specific issue type (if configured)

​

Priority Levels

• Highest: Critical issues blocking customers
• High: Important issues needing prompt attention
• Medium: Standard priority (default)
• Low: Minor issues or improvements
• Lowest: Nice-to-have enhancements

​

Assignee

• Use Jira username (not display name)
• Leave empty to use project default assignment
• Can be overridden by Jira automation rules

​

Labels

• ai-escalation: Identifies AI-escalated issues
• customer-reported: Originated from customer
• channel-livechat, channel-sms: Indicates channel
• urgent, billing, technical: Topic categories

​

Components

1. Find component names in Jira:
   • Go to Project Settings → Components
   • Note component names (case-sensitive)
2. Enter component names in escalation settings

• API
• Web Application
• Mobile App
• Billing System

​

Issue Content

• Escalation Summary: Brief overview
• Channel: Where conversation took place
• Full Transcript: Complete conversation history
• Customer Information: Name, email, identifier
• Reproduction Steps: If applicable
• Timestamp: When issue was reported
• Link to Conversation: Direct link in Support Unicorn

​

Escalation Triggers

​

Manual Only

• Customer explicitly reports a bug or problem
• AI determines it cannot assist and issue needs tracking

​

Message Count Trigger

​

Keyword Trigger

• Technical: “error”, “broken”, “not working”, “crash”
• Bug-related: “bug”, “issue”, “problem”, “glitch”
• Data: “missing”, “incorrect”, “wrong”

​

How It Works

1. Issue Created: New Jira issue created in specified project
2. AI Pauses: AI agent stops responding
3. Team Notified: Assignee and watchers receive Jira notification
4. Context Included: Full conversation in issue description
5. Workflow Started: Issue follows your Jira workflow
6. Updates Tracked: Status changes tracked in Jira

​

Managing Jira Escalations

​

Test Connection

1. Navigate to Agents → Your Agent → Escalations
2. Find your Jira escalation path
3. Click “Test Connection”
4. Verify test issue appears in Jira project

​

Activate/Deactivate

​

Edit Settings

1. Navigate to Agents → Your Agent → Escalations
2. Click the Jira escalation path
3. Update configuration
4. Click “Save Changes”

​

Best Practices

1. Consistent Labeling: Use standard labels for reporting
2. Appropriate Issue Types: Match escalation type to issue type
3. Component Mapping: Route to correct engineering teams
4. Clear Descriptions: Include reproduction steps when possible
5. SLA Tracking: Monitor resolution times for customer-reported issues
6. Feedback Loop: Update AI knowledge base from resolved issues

​

Example Use Cases

​

Use Case 1: Bug Reports

• Project: BUG
• Issue Type: Bug
• Priority: High
• Labels: ai-escalation, customer-reported, bug
• Components: API, Web Application
• Trigger: Keywords (“error”, “broken”, “crash”, “bug”)

​

Use Case 2: Feature Requests

• Project: PRODUCT
• Issue Type: Story
• Priority: Medium
• Labels: ai-escalation, feature-request, customer-feedback
• Assignee: Product Manager
• Trigger: Keywords (“feature”, “enhancement”, “would be great if”)

​

Use Case 3: Technical Support Issues

• Project: SUP
• Issue Type: Task
• Priority: Normal
• Labels: ai-escalation, technical-support
• Components: Customer Support
• Trigger: Message count > 7

​

Using in Procedures

For technical issues:
{{escalate:escal_jira_bugs}}
I've created a bug report for our engineering team. We'll investigate and update you soon.

​

Troubleshooting

​

Issues Not Being Created

• Verify Jira integration is connected and active
• Check API token has issue creation permissions
• Ensure project key is correct (case-sensitive)
• Verify escalation path is activated
• Test connection using “Test Connection” button

​

”Project Not Found” Error

• Verify project key is exactly correct
• Check that API user has access to the project
• Ensure project exists and is active

​

Invalid Issue Type

• Verify issue type is available in the project
• Check spelling and case of issue type name
• Ensure issue type is not archived or disabled

​

Assignee Not Set

• Verify username is correct (not display name)
• Check that user has access to the project
• Ensure user is active in Jira
• Leave empty to use project defaults

​

Labels or Components Not Applied

• Verify exact names (case-sensitive)
• Check that components exist in project
• Ensure API user has permission to apply labels

​

Advanced Configuration

​

Custom Fields

• Map conversation data to Jira custom fields
• Support for text, select, and user picker fields
• Useful for tracking customer metadata

​

Workflow Automation

• Auto-transition issues based on status
• Send notifications to specific channels
• Link related issues automatically
• Update issue fields based on conditions

​

Next Steps