Understanding Tickets

StackPilot's ticket system is the central hub for incident management and investigation. When issues are detected, StackPilot automatically creates tickets that track the entire lifecycle from initial alert to final resolution.

Ticket Lifecycle

1. Ticket Creation

Tickets are created automatically when:

• Alerts are triggered from connected monitoring tools (Sentry, Datadog, New Relic)
• CI/CD failures are detected from deployment pipelines
• Manual incidents are reported by team members
• Anomalies are detected by StackPilot's AI analysis

2. Automatic Triage

Upon creation, StackPilot immediately:

• Assigns severity levels based on error type and impact
• Correlates with recent deployments and code changes
• Gathers relevant logs and traces from connected systems
• Identifies similar past incidents for pattern recognition

3. Investigation States

Tickets progress through these states:

Open

• Initial state when ticket is created
• Basic triage information is available
• Waiting for AI investigation to begin

Need Info

• StackPilot requires additional information or access
• May need configuration updates or permission changes
• Manual input may be required from team members

Investigating

• AI agent is actively analyzing the incident
• Correlating data across connected systems
• Building investigation timeline and evidence

Investigated

• Analysis is complete with findings available
• Root cause hypothesis has been generated
• Code fix recommendations may be available

Code Generated

• StackPilot has generated a proposed code fix
• Pull request may be automatically created
• Ready for developer review and testing

Resolved

• Issue has been fixed and verified
• Incident is closed with full documentation
• Post-mortem and learnings are captured

Ticket Information

Basic Details

Each ticket includes:

• Title - Descriptive summary of the issue
• Description - Detailed incident information
• Severity - Impact level (Low, Medium, High, Critical)
• Status - Current investigation state
• Created/Updated - Timestamps for tracking
• Assigned Team - Responsible team or individual

Technical Data

StackPilot automatically gathers:

• Stack traces and error messages
• Related code changes and recent deployments
• Metrics and logs from the time of incident
• Similar incident history and patterns
• Affected services and dependencies

Investigation Timeline

Each ticket maintains a chronological record of:

• Initial alert and detection
• AI analysis steps and findings
• Code correlations and deployment tracking
• Manual interventions by team members
• Resolution actions and verification

Ticket Management

Viewing Tickets

Access tickets through:

• Project Dashboard - Overview of all project tickets
• Ticket Filters - Filter by status, severity, or time period
• Search - Find specific tickets by keywords or error messages
• Team Views - See tickets assigned to your team

Manual Ticket Creation

Create tickets manually when:

1. Click "Create Ticket" in the project dashboard
2. Provide incident description and relevant details
3. Set severity level and initial status
4. Add relevant links or additional context
5. StackPilot will begin automatic investigation

Ticket Actions

Available actions depend on ticket state:

• Add comments and manual findings
• Change severity if impact assessment changes
• Assign to team members for specific expertise
• Link related tickets for complex incidents
• Export data for external tools or reporting

Working with AI Investigation

Understanding AI Analysis

StackPilot's AI investigates by:

• Code correlation - Linking errors to recent code changes
• Pattern matching - Comparing with similar past incidents
• Dependency analysis - Understanding service interactions
• Temporal correlation - Analyzing timing of events and deployments

AI-Generated Insights

Look for these AI-powered features:

• Root cause hypothesis with confidence levels
• Suspected code commits that may have caused the issue
• Recommended investigation steps for manual follow-up
• Similar incident references with resolution patterns
• Automated fix suggestions when patterns are clear

Collaborating with AI

Best practices for working with StackPilot:

• Review AI findings but apply engineering judgment
• Add manual observations to enhance AI learning
• Validate proposed fixes before applying to production
• Provide feedback on AI accuracy to improve future analysis

Ticket Analytics

Metrics Tracking

StackPilot tracks key incident metrics:

• Mean Time to Detection (MTTD) - How quickly issues are found
• Mean Time to Resolution (MTTR) - Time from detection to fix
• AI Investigation Rate - Percentage of tickets investigated by AI
• Resolution Success Rate - Accuracy of AI-proposed fixes

Team Performance

Understanding team incident response:

• Ticket volume trends over time
• Resolution time patterns by severity
• Most common incident types and root causes
• AI assistance effectiveness for different issue categories

Continuous Improvement

Use ticket data to:

• Identify recurring issues that need architectural fixes
• Improve monitoring and alerting accuracy
• Optimize AI investigation through better integrations
• Build team playbooks from successful resolution patterns

Best Practices

Effective Ticket Management

• Review tickets promptly when they're created
• Add context when AI analysis seems incomplete
• Validate AI findings with manual investigation when needed
• Document manual steps to help AI learn for future incidents

Team Collaboration

• Use ticket comments for team communication during incidents
• Share investigation findings to build collective knowledge
• Tag relevant team members for specific expertise areas
• Update ticket status as investigation progresses

Integration Optimization

• Ensure complete tool integration for comprehensive data
• Configure proper alerting thresholds to avoid noise
• Maintain up-to-date team assignments for proper routing
• Regular review connection health for reliable data flow

By understanding how tickets work in StackPilot, your team can leverage AI-powered incident investigation while maintaining the human oversight necessary for complex problem-solving.