Workflow development

Human in the Loop

Add human oversight and approval steps to your automated workflows

Human-in-the-loop (HITL) enables you to incorporate human judgment and approval at critical decision points in your workflows.

Why Human Oversight?

  • Compliance Requirements - Meet regulatory approval requirements
  • Quality Assurance - Verify AI decisions before execution
  • Risk Management - Add checkpoints for high-stakes actions
  • Exception Handling - Route edge cases to human experts

HITL Node Types

Approval Node

Pause workflow execution until a designated approver takes action:

  • Single approver or approval chains
  • Timeout and escalation rules
  • Approval with comments

Review Node

Allow humans to review and modify data before proceeding:

  • Data validation checkpoints
  • Content review and editing
  • Quality scoring

Decision Node

Present options for human decision-making:

  • Multiple choice routing
  • Conditional branching based on human input
  • Feedback collection

Configuration Options

  • Assignees - Define who can approve or review
  • Notifications - Email, Slack, or webhook alerts
  • Timeouts - Auto-escalate or auto-approve rules
  • Audit Trail - Track all human decisions

Best Practices

  1. Strategic Placement - Add HITL at high-impact decision points
  2. Clear Instructions - Provide context for human reviewers
  3. Escalation Paths - Define backup approvers
  4. Performance Tracking - Monitor approval times and bottlenecks

Ask AI

FlowGenX Documentation

How can I help you?

Ask me anything about FlowGenX AI - workflows, agents, integrations, and more.

AI responses based on FlowGenX docs