Workflow Examples

Learn from practical workflow examples that demonstrate common patterns and best practices. These examples cover various use cases from simple automation to complex AI-powered pipelines.

Example 1: Document Processing Pipeline

Use Case: Automatically extract information from uploaded invoices and store in database.

Workflow Design

Form Trigger (Invoice upload)
  → PDF Parser (Extract text and tables)
  → Entity Extraction (Find invoice details)
  → AI Gateway (Validate and categorize)
  → Conditional (Validation check)
    → [Valid] → MongoDB (Store invoice)
              → Gmail (Send confirmation)
    → [Invalid] → Gmail (Request clarification)
  → API Return (Confirmation)

Configuration Details

Form Trigger:

{
  "fields": [
    {
      "name": "invoice_file",
      "type": "file",
      "accept": ".pdf",
      "maxSize": "10MB",
      "required": true
    },
    {
      "name": "customer_email",
      "type": "email",
      "required": true
    }
  ]
}

PDF Parser:

{
  "file": "{{ trigger.formData.invoice_file }}",
  "extractTables": true,
  "extractText": true
}

Entity Extraction:

{
  "text": "{{ pdfParser.text }}",
  "entityTypes": [
    "INVOICE_NUMBER",
    "DATE",
    "MONEY",
    "ORGANIZATION"
  ]
}

AI Gateway (Validation):

{
  "model": "gpt-4",
  "systemPrompt": "Validate invoice data and categorize expense type.",
  "userPrompt": `
    Invoice Data:
    ${entityExtraction.entities}

    Validate and return JSON:
    {
      "isValid": boolean,
      "category": string,
      "confidence": number
    }
  `,
  "temperature": 0.3
}

MongoDB:

{
  "action": "insert_one",
  "collection": "invoices",
  "document": {
    "invoiceNumber": "{{ entityExtraction.entities.INVOICE_NUMBER }}",
    "date": "{{ entityExtraction.entities.DATE }}",
    "amount": "{{ entityExtraction.entities.MONEY }}",
    "vendor": "{{ entityExtraction.entities.ORGANIZATION }}",
    "category": "{{ aiGateway.category }}",
    "uploadedAt": "{{ execution.timestamp }}",
    "customerEmail": "{{ trigger.formData.customer_email }}"
  }
}

Benefits:

Automated invoice processing
AI-powered validation
Structured data extraction
Error handling and notifications

Example 2: Customer Support Automation

Use Case: Automatically triage and respond to customer support emails.

Workflow Design

Email Trigger (Support inbox)
  → Email Parser (Extract content)
  → Document Classification (Categorize issue)
  → Conversation Memory (Load customer history)
  → Conditional (Priority routing)
    → [Urgent] → AI Gateway (Immediate response)
              → Gmail (Send response)
              → Slack (Alert team)
    → [Normal] → AI Gateway (Draft response)
              → MongoDB (Queue for review)
    → [FAQ] → Knowledge Base (Find answer)
            → AI Gateway (Format response)
            → Gmail (Auto-respond)

Configuration Details

Document Classification:

{
  "text": "{{ emailParser.body }}",
  "categories": [
    "urgent_bug",
    "feature_request",
    "billing_question",
    "faq",
    "other"
  ],
  "multiLabel": false
}

Conversation Memory:

{
  "sessionId": "{{ emailParser.from }}",
  "operation": "load",
  "maxMessages": 5
}

Knowledge Base (FAQ path):

{
  "collection": "support_articles",
  "query": "{{ emailParser.subject }} {{ emailParser.body }}",
  "topK": 3,
  "threshold": 0.75
}

AI Gateway (Auto-response):

{
  "model": "claude-3-sonnet",
  "systemPrompt": "You are a helpful customer support agent. Provide clear, friendly responses.",
  "userPrompt": `
    Customer Question:
    ${emailParser.body}

    Relevant Article:
    ${knowledgeBase.results[0].content}

    Previous Conversation:
    ${conversationMemory.history}

    Write a helpful response.
  `,
  "temperature": 0.7
}

Benefits:

24/7 automated support
Intelligent prioritization
Contextual responses
Team alerts for urgent issues

Example 3: GitHub Issue Automation

Use Case: Automatically create GitHub issues from form submissions with AI-powered analysis.

Workflow Design

Form Trigger (Bug report)
  → Entity Extraction (Parse bug details)
  → AI Gateway (Analyze severity and categorize)
  → GitHub MCP (Search for duplicates)
  → Conditional (Duplicate check)
    → [Not Duplicate] → GitHub MCP (Create issue)
                      → Slack MCP (Notify team)
                      → MongoDB (Track submission)
    → [Duplicate] → GitHub MCP (Add comment to existing)
                  → Gmail (Notify reporter)
  → API Return (Issue URL)

Configuration Details

Form Trigger:

{
  "fields": [
    {
      "name": "title",
      "type": "text",
      "required": true
    },
    {
      "name": "description",
      "type": "textarea",
      "required": true
    },
    {
      "name": "steps_to_reproduce",
      "type": "textarea",
      "required": true
    },
    {
      "name": "environment",
      "type": "select",
      "options": ["Production", "Staging", "Development"]
    }
  ]
}

AI Gateway (Analysis):

{
  "model": "gpt-4",
  "systemPrompt": "Analyze bug reports and determine severity and category.",
  "userPrompt": `
    Bug Report:
    Title: ${trigger.formData.title}
    Description: ${trigger.formData.description}
    Steps: ${trigger.formData.steps_to_reproduce}
    Environment: ${trigger.formData.environment}

    Return JSON:
    {
      "severity": "critical|high|medium|low",
      "category": "bug|performance|security|ui|other",
      "labels": ["array", "of", "labels"]
    }
  `,
  "responseFormat": "json_object"
}

GitHub MCP (Search):

{
  "action": "search_issues",
  "repo": "company/product",
  "query": "{{ trigger.formData.title }}",
  "state": "open",
  "limit": 5
}

GitHub MCP (Create Issue):

{
  "action": "create_issue",
  "repo": "company/product",
  "title": "{{ trigger.formData.title }}",
  "body": `
## Description
${trigger.formData.description}

## Steps to Reproduce
${trigger.formData.steps_to_reproduce}

## Environment
${trigger.formData.environment}

## Analysis
Severity: ${aiGateway.severity}
Category: ${aiGateway.category}
  `,
  "labels": "{{ aiGateway.labels }}",
  "assignees": "{{ aiGateway.severity === 'critical' ? ['on-call-engineer'] : [] }}"
}

Benefits:

Automated issue creation
AI-powered analysis
Duplicate detection
Intelligent assignment

Example 4: Daily Report Generation

Use Case: Generate and email daily business metrics report.

Workflow Design

Schedule Trigger (Daily at 8 AM)
  → MongoDB (Fetch yesterday's orders)
  → Loop (Calculate metrics)
    → Map (Transform data)
  → PostgreSQL (Get user analytics)
  → Merge (Combine data sources)
  → AI Gateway (Generate insights)
  → Excel Generator (Create report)
  → Gmail (Send to stakeholders)
  → Amazon S3 (Archive report)

Configuration Details

Schedule Trigger:

Schedule: 0 8 * * 1-5
Timezone: America/New_York
Description: Weekdays at 8 AM

MongoDB:

{
  "action": "find",
  "collection": "orders",
  "filter": {
    "createdAt": {
      "$gte": "{{ new Date().setDate(new Date().getDate() - 1) }}",
      "$lt": "{{ new Date() }}"
    }
  }
}

Map (Calculate Metrics):

{
  "input": "{{ mongoOrders.results }}",
  "transform": {
    "totalRevenue": "{{ items.reduce((sum, order) => sum + order.amount, 0) }}",
    "orderCount": "{{ items.length }}",
    "avgOrderValue": "{{ totalRevenue / orderCount }}",
    "topProducts": "{{ /* group and sort by product */ }}"
  }
}

AI Gateway (Insights):

{
  "model": "gpt-4",
  "systemPrompt": "Analyze business metrics and provide actionable insights.",
  "userPrompt": `
    Yesterday's Metrics:
    - Revenue: $${metrics.totalRevenue}
    - Orders: ${metrics.orderCount}
    - Avg Order Value: $${metrics.avgOrderValue}
    - Top Products: ${metrics.topProducts}

    User Analytics:
    ${postgresData.analytics}

    Provide 3-5 key insights and recommendations.
  `,
  "temperature": 0.5
}

Gmail:

{
  "to": ["team@example.com"],
  "subject": "Daily Business Report - {{ new Date().toLocaleDateString() }}",
  "body": `
    <h1>Daily Report</h1>

    <h2>Key Metrics</h2>
    <ul>
      <li>Revenue: $${metrics.totalRevenue}</li>
      <li>Orders: ${metrics.orderCount}</li>
      <li>Avg Order: $${metrics.avgOrderValue}</li>
    </ul>

    <h2>AI Insights</h2>
    ${aiGateway.insights}

    <p>Full report attached.</p>
  `,
  "attachments": ["{{ excelReport.file }}"]
}

Benefits:

Automated reporting
Multi-source data aggregation
AI-generated insights
Consistent delivery schedule

Example 5: Content Moderation Pipeline

Use Case: Automatically moderate user-generated content using AI.

Workflow Design

REST API Trigger (Content submission)
  → AI Gateway (Content analysis)
  → Conditional (Moderation result)
    → [Approved] → MongoDB (Publish content)
                 → API Return (200 OK)
    → [Flagged] → AI Gateway (Detailed review)
                → Conditional (Severity)
                  → [Minor] → MongoDB (Queue for review)
                  → [Severe] → MongoDB (Reject)
                            → Gmail (Notify user)
                → API Return (202 Pending)
    → [Rejected] → MongoDB (Log rejection)
                 → API Return (400 Rejected)

Configuration Details

AI Gateway (Content Analysis):

{
  "model": "claude-3-sonnet",
  "systemPrompt": `Analyze content for policy violations. Check for:
    - Hate speech or harassment
    - Violence or threats
    - Adult content
    - Spam or scams
    - Misinformation
  `,
  "userPrompt": `
    Content to analyze:
    ${trigger.body.content}

    Return JSON:
    {
      "decision": "approved|flagged|rejected",
      "violations": ["list of issues"],
      "confidence": 0-1,
      "explanation": "brief explanation"
    }
  `,
  "responseFormat": "json_object",
  "temperature": 0.2
}

AI Gateway (Detailed Review):

{
  "model": "gpt-4",
  "systemPrompt": "Provide detailed content moderation analysis.",
  "userPrompt": `
    Content: ${trigger.body.content}
    Initial Analysis: ${contentAnalysis.explanation}
    Violations: ${contentAnalysis.violations}

    Provide:
    1. Severity (minor|moderate|severe)
    2. Specific policy violations
    3. Recommended action
    4. User feedback message
  `,
  "temperature": 0.3
}

Benefits:

Real-time content moderation
Multi-tier review process
Detailed violation tracking
User feedback

Example 6: E-Commerce Order Processing

Use Case: Process e-commerce orders with payment, inventory, and fulfillment.

Workflow Design

Webhook Trigger (New order)
  → Stripe MCP (Process payment)
  → Conditional (Payment status)
    → [Success] → MongoDB (Update inventory)
                → Conditional (Stock available)
                  → [Yes] → Notion MCP (Create fulfillment task)
                          → Webflow MCP (Update order status)
                          → Gmail (Order confirmation)
                  → [No] → MongoDB (Backorder)
                         → Gmail (Backorder notification)
    → [Failed] → MongoDB (Log failure)
              → Gmail (Payment failed notification)

Configuration Details

Stripe MCP:

{
  "action": "create_payment_intent",
  "amount": "{{ trigger.body.amount }}",
  "currency": "usd",
  "customer": "{{ trigger.body.customerId }}",
  "metadata": {
    "orderId": "{{ trigger.body.orderId }}",
    "items": "{{ trigger.body.items }}"
  }
}

MongoDB (Update Inventory):

{
  "action": "update_many",
  "collection": "products",
  "operations": "{{ trigger.body.items.map(item => ({
    filter: { _id: item.productId },
    update: { $inc: { stock: -item.quantity } }
  })) }}"
}

Notion MCP:

{
  "action": "create_page",
  "database": "fulfillment_queue",
  "properties": {
    "Order ID": "{{ trigger.body.orderId }}",
    "Customer": "{{ trigger.body.customerName }}",
    "Items": "{{ trigger.body.items }}",
    "Status": "Pending",
    "Priority": "{{ trigger.body.shipping === 'express' ? 'High' : 'Normal' }}"
  }
}

Benefits:

Automated payment processing
Real-time inventory management
Integrated fulfillment tracking
Customer notifications

Common Workflow Patterns

Pattern: Event-Driven Integration

External Event (Webhook)
  → Validate & Parse
  → Business Logic
  → Update Systems
  → Notify Stakeholders

Use Cases:

Payment webhooks
Git repository events
Third-party integrations

Pattern: Scheduled Batch Processing

Schedule Trigger
  → Fetch Data (Multiple sources)
  → Merge & Transform
  → Process (Loop/Map)
  → Store Results
  → Generate Report

Use Cases:

Daily reports
Data synchronization
Batch analytics

Pattern: AI Pipeline

Input Trigger
  → Data Preparation
  → AI Model (Analysis/Generation)
  → Validation
  → Post-Processing
  → Output

Use Cases:

Content generation
Data extraction
Sentiment analysis
Classification

Pattern: Multi-System Orchestration

API Request
  → Fetch from System A
  → Fetch from System B (Parallel)
  → Merge Data
  → Transform
  → Update System C
  → Return Response

Use Cases:

Microservice orchestration
Data aggregation
Cross-system updates

Best Practices from Examples

Error Handling

Always use Conditional nodes for validation
Provide user-friendly error messages
Log failures for debugging
Implement retry logic for transient failures

Performance

Use parallel execution where possible (Merge node)
Batch operations in loops
Cache frequently accessed data
Set appropriate timeouts

Security

Validate all inputs
Use HMAC for webhooks
Don't expose sensitive data in responses
Use Data Sources for credentials

Maintainability

Use descriptive node names
Add notes to complex logic
Version workflows properly
Document expected data formats

Next Steps

Start Simple

Begin with a simple workflow and gradually add complexity. Test thoroughly at each step before adding more nodes.

Example 1: Document Processing Pipeline​

Workflow Design​

Configuration Details​

Example 2: Customer Support Automation​

Workflow Design​

Configuration Details​

Example 3: GitHub Issue Automation​

Workflow Design​

Configuration Details​

Example 4: Daily Report Generation​

Workflow Design​

Configuration Details​

Example 5: Content Moderation Pipeline​

Workflow Design​

Configuration Details​

Example 6: E-Commerce Order Processing​

Workflow Design​

Configuration Details​

Common Workflow Patterns​

Pattern: Event-Driven Integration​

Pattern: Scheduled Batch Processing​

Pattern: AI Pipeline​

Pattern: Multi-System Orchestration​

Best Practices from Examples​

Error Handling​

Performance​

Security​

Maintainability​

Next Steps​

Example 1: Document Processing Pipeline

Workflow Design

Configuration Details

Example 2: Customer Support Automation

Workflow Design

Configuration Details

Example 3: GitHub Issue Automation

Workflow Design

Configuration Details

Example 4: Daily Report Generation

Workflow Design

Configuration Details

Example 5: Content Moderation Pipeline

Workflow Design

Configuration Details

Example 6: E-Commerce Order Processing

Workflow Design

Configuration Details

Common Workflow Patterns

Pattern: Event-Driven Integration

Pattern: Scheduled Batch Processing

Pattern: AI Pipeline

Pattern: Multi-System Orchestration

Best Practices from Examples

Error Handling

Performance

Security

Maintainability

Next Steps