Skip to main content

AI Configuration

Configure AI behavior, limits, and settings for conversational forms and AI agents.

AI Settings​

Access AI settings from Settings > AI Configuration.

Global AI Settings​

API Configuration​

  • OpenAI API Key: Your OpenAI API key for AI features
  • Model Selection: Choose AI model (GPT-4, GPT-3.5, etc.)
  • Temperature: Control randomness (0-1, default 0.7)
  • Max Tokens: Maximum response length

Default Persona​

Set default persona for new conversational forms:

  • Style: Professional, Friendly, Casual, Empathetic
  • Default Tone: Default tone description
  • Default Behaviors: Default behavior list
  • Default Restrictions: Default restrictions

Conversation Defaults​

  • Default Max Turns: Default maximum conversation exchanges
  • Default Max Duration: Default maximum conversation time (minutes)
  • Default Min Confidence: Default minimum confidence threshold (0-1)

Conversational Form Configuration​

Per-Form Settings​

Each conversational form has its own configuration:

Objective​

Clear statement of what the form should accomplish:

"Collect IT support ticket information including issue 
description, urgency level, device information, and 
user contact details."

Context​

Background information for the AI:

"We're an IT support team for a 500-person company. 
We handle hardware issues, software problems, network 
connectivity, and account access requests."

Topics​

Define conversation topics:

[
  {
    id: "issue-description",
    name: "Issue Description",
    description: "Detailed description of the technical problem",
    priority: "required",
    depth: "deep"
  },
  {
    id: "urgency-level",
    name: "Urgency Level",
    description: "How urgent is this issue?",
    priority: "required",
    depth: "surface"
  }
]

Priority Options:

  • required: Must be collected
  • important: Should be collected if possible
  • optional: Collect if relevant

Depth Options:

  • surface: Basic information only
  • moderate: Some detail expected
  • deep: Comprehensive information

Persona​

Customize AI personality:

{
  style: "professional",
  tone: "Helpful and patient, but efficient",
  behaviors: [
    "Ask one question at a time",
    "Provide reassurance when appropriate",
    "Clarify technical terms if needed"
  ],
  restrictions: [
    "Don't make technical diagnoses",
    "Don't promise specific resolution times",
    "Don't ask for passwords or sensitive credentials"
  ]
}

Conversation Limits​

{
  maxTurns: 15,        // Maximum conversation exchanges
  maxDuration: 10,     // Minutes
  minConfidence: 0.8    // 0-1 threshold
}

Extraction Schema​

Define structured data fields:

[
  {
    id: "issueCategory",
    name: "Issue Category",
    type: "select",
    required: true,
    options: ["hardware", "software", "network", "other"],
    validation: {
      required: true
    }
  },
  {
    id: "description",
    name: "Description",
    type: "textarea",
    required: true,
    validation: {
      minLength: 10,
      maxLength: 1000
    }
  }
]

Advanced Configuration​

Custom Prompts​

Override default prompt templates:

{
  greeting: "Hi! I'm here to help you...",
  followUp: "Can you tell me more about {topic}?",
  clarification: "I want to make sure I understand...",
  confirmation: "Great! I've collected: {summary}...",
  completion: "Thank you! Your ticket has been submitted..."
}

Topic Dependencies​

Define relationships between topics:

{
  "deviceInfo": {
    dependsOn: ["issueCategory"],
    condition: "issueCategory === 'hardware'"
  }
}

Confidence Thresholds​

Set minimum confidence per field:

{
  "issueCategory": 0.9,    // High confidence required
  "description": 0.7,     // Medium confidence OK
  "deviceInfo": 0.5       // Low confidence acceptable
}

AI Agent Configuration​

Agent Settings​

Configure individual agents:

Form Generator​

  • Default Field Types: Preferred field types
  • Auto-Validation: Auto-add validation rules
  • Multi-Page Threshold: When to suggest multi-page

Inline Suggestions​

  • Suggestion Frequency: How often to show (always, sometimes, rarely)
  • Confidence Threshold: Minimum confidence for suggestions
  • Learning Mode: Learn from your patterns

Formula Assistant​

  • Preferred Syntax: Formula syntax style
  • Auto-Optimize: Auto-optimize formulas
  • Error Detection: Level of error checking

Form Optimization​

  • Optimization Level: Aggressive, Moderate, Conservative
  • Focus Areas: Performance, UX, Accessibility, All
  • Auto-Apply: Auto-apply safe optimizations

Usage Limits​

Tier-Based Limits​

AI usage is limited by subscription tier:

TierAI Generations/Month
Free10
Pro100
Team500
EnterpriseUnlimited

Monitoring Usage​

Track AI usage:

  • Current Usage: Generations used this month
  • Remaining: Generations remaining
  • Usage History: Historical usage data
  • Per-Agent Usage: Usage by agent type

Usage Alerts​

Configure alerts:

  • Warning Threshold: Alert when X% used
  • Limit Reached: Alert when limit reached
  • Reset Notification: Alert when limit resets

Best Practices​

  1. Clear Objectives: Write clear, specific objectives
  2. Relevant Context: Provide sufficient business context
  3. Appropriate Limits: Set reasonable conversation limits
  4. Test Configuration: Test before publishing
  5. Monitor Usage: Track AI usage regularly
  6. Iterate: Improve based on results

Troubleshooting​

Low Confidence Scores​

  • Provide more context
  • Clarify topic descriptions
  • Adjust confidence thresholds
  • Review extraction schema

Conversations Too Long​

  • Reduce max turns
  • Simplify topics
  • Adjust depth levels
  • Improve topic descriptions

Poor Extraction​

  • Review extraction schema
  • Improve topic descriptions
  • Adjust confidence thresholds
  • Test with sample conversations

Next Steps​