AI Agents

NetPad includes 15+ AI agents to help with form building, optimization, data processing, compliance, and RAG (Retrieval-Augmented Generation). These agents use AI to automate tasks and provide intelligent assistance.

Agent Categories

AI agents are organized into five categories:

Form Building Agents - Help create and configure forms
Form Enhancement Agents - Improve existing forms
Data Processing Agents - Process and analyze data
Compliance & Quality Agents - Ensure quality and compliance
RAG Agents - Knowledge-guided conversational forms with document grounding

Form Building Agents

Form Generator

Tier: Pro+
Description: Create complete forms from natural language descriptions

Usage:

Describe what you want in natural language
AI generates form structure
Review and customize generated form
Publish

Example:

Input: "Create a contact form with name, email, phone, 
        message, and a checkbox for newsletter subscription"

Output: Generated form with all specified fields, 
        validation rules, and basic styling

Features:

Generates field types automatically
Adds appropriate validation
Suggests conditional logic
Creates multi-page forms when needed

Field Type Detection

Tier: Free
Description: Auto-detect field types from sample data

Usage:

Provide sample data (CSV, JSON, or MongoDB documents)
AI analyzes data patterns
Suggests appropriate field types
Apply suggestions to form

Example:

Input: Sample data with dates, emails, numbers

Output: Date fields → Date Picker
        Email strings → Email Field
        Numbers → Number Field

Features:

Pattern recognition
Data type inference
Validation rule suggestions
Format detection

Conditional Logic Generator

Tier: Pro+
Description: Auto-generate show/hide rules based on form structure

Usage:

Describe desired conditional behavior
AI generates conditional logic rules
Review and test rules
Apply to form

Example:

Input: "Show 'Company Name' field only if 'Account Type' 
        is 'Business'"

Output: Conditional rule:
        IF accountType === 'business' 
        THEN show companyName

Features:

Natural language to rules
Complex condition support
Validation of logic
Testing suggestions

Validation Pattern Generator

Tier: Pro+
Description: Generate regex patterns for validation

Usage:

Describe validation requirements
AI generates regex pattern
Test pattern
Apply to field

Example:

Input: "Phone number in format (XXX) XXX-XXXX"

Output: /^\(\d{3}\) \d{3}-\d{4}$/

Features:

Pattern generation
Pattern explanation
Test cases
Edge case handling

Form Enhancement Agents

Inline Suggestions

Tier: Pro+
Description: Real-time field suggestions while building forms

Usage:

Works automatically while building
Suggests fields based on context
Recommends validation rules
Suggests conditional logic

Features:

Context-aware suggestions
Real-time assistance
Learning from your patterns
Non-intrusive UI

Formula Assistant

Tier: Pro+
Description: Help build computed field formulas

Usage:

Describe what you want to calculate
AI suggests formula syntax
Test formula
Apply to computed field

Example:

Input: "Calculate total price as quantity times unit price, 
        plus 10% tax"

Output: (quantity * unitPrice) * 1.1

Features:

Natural language to formulas
Syntax validation
Error detection
Optimization suggestions

Form Optimization

Tier: Team+
Description: Analyze and improve forms

Usage:

Run optimization analysis
Review suggestions
Apply improvements
Test optimized form

Suggestions Include:

Field order optimization
Validation improvements
Conditional logic simplification
Multi-page form suggestions
Performance optimizations

Features:

Performance analysis
UX recommendations
Accessibility improvements
Mobile optimization

Data Processing Agents

Response Processing

Tier: Team+
Description: Process and transform form responses

Usage:

Select responses to process
Describe transformation needed
AI processes responses
Review transformed data

Example:

Input: "Normalize phone numbers to format XXX-XXX-XXXX"

Output: All phone numbers transformed to consistent format

Features:

Data normalization
Format conversion
Data enrichment
Bulk processing

Response Insights

Tier: Team+
Description: Analyze submission patterns and trends

Usage:

Select time range
Run insights analysis
Review insights
Export insights report

Insights Include:

Submission trends
Field completion patterns
Common validation errors
User behavior patterns
Anomaly detection

Features:

Trend analysis
Pattern recognition
Anomaly detection
Predictive insights

Auto-Translation

Tier: Team+
Description: Translate forms to other languages

Usage:

Select target language(s)
AI translates form content
Review translations
Publish translated forms

Features:

Multi-language support
Context-aware translation
Field label translation
Validation message translation
Cultural adaptation

Compliance & Quality Agents

Compliance Audit

Tier: Team+
Description: Check forms for regulatory compliance

Usage:

Select compliance standards (HIPAA, GDPR, PCI-DSS, etc.)
Run audit
Review findings
Fix compliance issues

Checks Include:

Data encryption requirements
Consent mechanisms
Data retention policies
Access controls
Audit logging
Privacy policy links

Features:

Multi-standard support
Detailed findings
Remediation suggestions
Compliance scoring

Workflow Generator

Tier: Enterprise
Description: Generate workflows from natural language descriptions

Usage:

Describe workflow process
AI generates workflow structure
Review and customize
Activate workflow

Example:

Input: "When a form is submitted, send an email notification 
        to the team, then save to MongoDB, and if it's high 
        priority, also send a Slack message"

Output: Generated workflow with:
        - Form Trigger
        - Email Send node
        - MongoDB Write node
        - Conditional node
        - Slack node

Features:

Natural language to workflows
Node selection
Connection mapping
Error handling
Testing suggestions

RAG Agents

RAG (Retrieval-Augmented Generation) agents enable document-grounded conversational forms. Requirements vary by deployment mode:

Deployment Mode	Subscription Required	Infrastructure Required
Cloud (netpad.io)	Team or Enterprise	M10+ Atlas cluster
Self-Hosted	Any tier (including Free)	Atlas Local (Docker)

Knowledge-Guided Conversation

Tier: Team+ (Cloud) / Any (Self-Hosted) Requirements: M10+ Atlas cluster (Cloud) or Atlas Local (Self-Hosted) Description: RAG-powered conversational forms with document grounding

Usage:

Upload knowledge base documents (PDF, DOCX, TXT)
Enable RAG in conversational form settings
AI uses documents to answer questions
Responses include source citations

Features:

Document-grounded responses
Source citations with traceability
Semantic document search
Context-aware answers
Schema-aware suggestions

Document Upload for RAG

Tier: Team+ (Cloud) / Any (Self-Hosted) Requirements: M10+ Atlas cluster (Cloud) or Atlas Local (Self-Hosted) Description: Manage knowledge base documents for RAG

Usage:

Upload documents to form knowledge base
Add metadata (title, description, tags)
Documents automatically processed and indexed
Select documents for use in forms

Features:

Multi-format support (PDF, DOCX, TXT up to 5MB)
Automatic text extraction
Intelligent sentence-aware chunking
OpenAI embeddings (text-embedding-3-small)
MongoDB Atlas Vector Search indexing

Vector Search

Tier: Team+ (Cloud) / Any (Self-Hosted) Requirements: M10+ Atlas cluster (Cloud) or Atlas Local (Self-Hosted) Description: Semantic document retrieval using MongoDB Atlas Vector Search

Usage:

Automatically used when RAG is enabled
Searches document chunks by semantic similarity
Returns most relevant chunks for context

Features:

Semantic similarity search
Relevance scoring
Configurable retrieval parameters (maxChunks, minScore, retrievalThreshold)
Fast vector search performance

Using AI Agents

Accessing Agents

Agents are available in different contexts:

Form Builder: Inline suggestions, formula assistant
Form Settings: Form optimization, compliance audit
Analytics: Response insights
Workflow Builder: Workflow generator
Settings: Template management, bulk operations

Agent Usage Limits

AI agent usage is limited by subscription tier:

Tier	AI Generations/Month
Free	10
Pro	100
Team	500
Enterprise	Unlimited

Best Practices

Clear Descriptions: Be specific when describing what you want
Review Output: Always review AI-generated content
Test Thoroughly: Test AI suggestions before publishing
Iterate: Refine based on results
Monitor Usage: Track AI usage to stay within limits

Agent Configuration

Some agents can be configured:

Confidence Thresholds: Minimum confidence for suggestions
Suggestion Frequency: How often to show suggestions
Learning Mode: Learn from your patterns
Custom Prompts: Customize agent behavior

Next Steps

Conversational Forms - Create AI-powered forms
Knowledge-Guided Forms (RAG) - Add document-grounded AI
Template Management - Use templates
Configuration - Configure AI settings

Agent Categories​

Form Building Agents​

Form Generator​

Field Type Detection​

Conditional Logic Generator​

Validation Pattern Generator​

Form Enhancement Agents​

Inline Suggestions​

Formula Assistant​

Form Optimization​

Data Processing Agents​

Response Processing​

Response Insights​

Auto-Translation​

Compliance & Quality Agents​

Compliance Audit​

Workflow Generator​

RAG Agents​

Knowledge-Guided Conversation​

Document Upload for RAG​

Vector Search​

Using AI Agents​

Accessing Agents​

Agent Usage Limits​

Best Practices​

Agent Configuration​

Next Steps​

Agent Categories

Form Building Agents

Form Generator

Field Type Detection

Conditional Logic Generator

Validation Pattern Generator

Form Enhancement Agents

Inline Suggestions

Formula Assistant

Form Optimization

Data Processing Agents

Response Processing

Response Insights

Auto-Translation

Compliance & Quality Agents

Compliance Audit

Workflow Generator

RAG Agents

Knowledge-Guided Conversation

Document Upload for RAG

Vector Search

Using AI Agents

Accessing Agents

Agent Usage Limits

Best Practices

Agent Configuration

Next Steps