Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.encoreos.io/llms.txt

Use this file to discover all available pages before exploring further.

This guide explains how to use AI-assisted wizard creation to rapidly build and customize workflow wizards from natural language descriptions.

Overview

The AI Wizard Generator uses advanced language models to:
  • Create complete wizard structures from descriptions
  • Generate appropriate fields and validation
  • Configure branching logic automatically
  • Suggest optimal step organization

Accessing AI Generation

  1. Navigate to Platform → Wizards
  2. Click New Wizard or open an existing wizard in the Builder
  3. Click the Generate with AI button (✨ icon) in the toolbar

Writing Effective Descriptions

The quality of your description directly impacts the generated wizard. Here are tips for best results:

Be Specific About Fields

Vague: “Create an employee form” Specific: “Create an employee onboarding wizard that collects first name, last name, email, phone number, hire date, department, and job title”

Mention Conditions

If you need branching logic, describe the conditions: ✅ “Create an onboarding wizard. If the employee is full-time, show benefits selection. If part-time, skip to equipment request.”

Specify Field Types

When precision matters: ✅ “Collect date of birth (date picker), salary (currency), and department (dropdown with options: HR, Engineering, Sales, Marketing)“

Include Validation Requirements

✅ “Email must be a valid company email (@company.com). Phone is optional. All other fields are required.”

Example Prompts

Simple Form Wizard

Create a contact form wizard with:
- Personal info step: name, email, phone
- Message step: subject, message (long text), urgency (low/medium/high)
- Review step

Branching Wizard

Create a support ticket wizard:
1. Issue type: Bug, Feature Request, or Question
2. If Bug: collect steps to reproduce, expected behavior, actual behavior
3. If Feature: collect description, business justification
4. If Question: collect topic, question details
5. All paths end with a review step

Complex Workflow

Create an employee exit interview wizard for HR:
- Exit reason: Resignation, Retirement, Termination, Other
- If Resignation: last day, forwarding address, exit interview availability
- If Retirement: retirement date, benefits continuation preferences
- All: equipment return checklist, knowledge transfer status
- Review and acknowledgment step

Understanding the Preview

After generation, you’ll see a preview of the wizard:

Step List

  • Form Steps show field count and types
  • Branch Steps display conditions and targets
  • Review Steps are marked accordingly

Field Details

Click to expand each step and see:
  • Field names and types
  • Required indicators
  • Help text suggestions
  • Validation rules

Branch Logic

Branch steps show:
  • Each condition with its target
  • Default fallback path
  • Visual path indicators

Editing Generated Wizards

After reviewing the preview:

Accept & Create

  • Creates the wizard with the generated configuration
  • Opens in the Builder for final adjustments
  • Saves as a draft (unpublished)

Edit in Builder

  • Loads the generated steps into the Builder
  • Make changes before saving
  • Full access to all configuration options

Regenerate

  • Not happy with the result? Click Regenerate
  • Modify your description and try again
  • Each generation is independent

Generate Similar Feature

Already have a wizard and want a variant?
  1. Open an existing wizard in the Builder
  2. Click More → Generate Similar
  3. Describe the changes you want:
    • “Add an emergency contact section”
    • “Remove the benefits step and add a contractor agreement”
    • “Make this wizard for part-time employees”

Best Practices

Start Simple

  1. Generate a basic version first
  2. Review and test the output
  3. Iterate with modifications

Combine with Manual Editing

AI generation is a starting point:
  1. Generate the initial structure
  2. Refine field labels and help text
  3. Adjust validation rules
  4. Fine-tune branching conditions

Version Your Changes

  1. Generate a wizard
  2. Publish as v1.0
  3. Use “Generate Similar” for modifications
  4. Publish as v1.1
  5. Compare versions if needed

Limitations

What AI Can Generate

✅ Form steps with common field types ✅ Basic branching logic ✅ Review steps ✅ Standard validation rules ✅ Help text and placeholders

What Requires Manual Work

❌ Custom step components ❌ Complex multi-level branching ❌ Integration with external systems ❌ Advanced regex validation patterns ❌ Custom field types

Rate Limits

  • Requests are rate-limited per organization
  • If you hit limits, wait a moment and retry
  • Contact support for higher limits

Troubleshooting

Generation Takes Too Long

  • Simplify your description
  • Break into smaller wizards
  • Check your internet connection

Poor Quality Output

  • Add more detail to your description
  • Specify field types explicitly
  • Include example values
  • Mention validation requirements

Missing Fields

  • Explicitly list all required fields
  • Don’t assume common fields are included
  • Add them manually in the Builder

Wrong Field Types

  • Specify types in your description
  • Example: “email (email field)” or “hire_date (date picker)”
  • Edit in Builder after generation

Privacy & Security

AI wizard generation:
  • Does not store your descriptions
  • Does not include sample data in training
  • Does not transmit existing wizard data
  • Is organization-scoped - only you see your generations
Important: Never include real PII/PHI in your descriptions. Use generic field names like “SSN” rather than actual values.