This section presents concrete examples of Actions successfully implemented across different sectors, providing detailed configurations and real-world use cases to inspire your implementations.
Example 1: E-commerce - Order and returns management
Scenario
An online store wants to automate return request management and order status checking through the assistant.
Action: Order status check
Name: Check Order Status
Description: Use this Action when the user wants to verify the status of an existing order. Requires the order number or email associated with the order.
URL: https://api.mystore.com/orders/status
Method: GET
Authentication: API Key in header (X-API-Key)
Parameters:
- order_number (required): Order number
- customer_email (optional): Email for additional verification
Action: Return request
Name: Create Return Request
Description: Creates a return request for an existing order. Requires order number, return reason and product details.
URL: https://api.mystore.com/returns
Method: POST
Content-Type: application/json
Parameters:
- order_number (required): Original order number
- customer_email (required): Customer email
- reason (required): Return reason
- product_ids (required): Product IDs to return
- notes (optional): Additional notes
Example conversation
Customer: "Hi, I would like to check the status of my order #ORD-12345"
Assistant: "I'll check your order status immediately... Order #ORD-12345 was shipped yesterday and will arrive tomorrow by 6:00 PM. You received the tracking via email."
Customer: "Actually I changed my mind about one of the products, can I return it?"
Assistant: "Certainly! I can help you create a return request. Which product would you like to return and what's the reason?"
Customer: "The blue t-shirt, size M. I ordered the wrong size."
Assistant: "Perfect! I've created return request #RET-7890 for the blue t-shirt size M. You'll receive the shipping label via email within 30 minutes."
Example 2: Professional services - Consultation booking
Scenario
A law firm wants to automate consultation booking and initial client qualification.
Action: Check availability
Name: Check Lawyer Availability
Description: Checks available slots for legal consultations based on date and type of consultation requested.
URL: https://calendar.lawfirm.com/api/availability
Method: GET
Parameters:
- date (required): Desired date (YYYY-MM-DD)
- consultation_type (required): Consultation type (civil, criminal, corporate)
- duration (optional, default: 60): Duration in minutes
Action: Book consultation
Name: Book Legal Consultation
Description: Creates a booking for legal consultation after checking availability.
URL: https://calendar.lawfirm.com/api/bookings
Method: POST
Parameters:
- client_name (required): Client full name
- client_email (required): Client email
- client_phone (required): Client phone
- appointment_date (required): Appointment date and time
- consultation_type (required): Type of consultation
- case_description (required): Brief case description
- urgency (optional): Urgency level (normal, high, urgent)
Example conversation
Client: "Good morning, I would need a consultation for an employment law matter"
Assistant: "Good morning! I'll be happy to help you book a consultation. When would you like the appointment?"
Client: "Possibly this week, as soon as possible"
Assistant: "I've checked availability: I can offer you Thursday March 15th at 2:30 PM or Friday March 16th at 10:00 AM. Which do you prefer?"
Client: "Thursday at 2:30 PM works perfectly"
Assistant: "Perfect! To complete the booking I need some information... I've booked your consultation for Thursday March 15th at 2:30 PM. You'll receive a confirmation email with all the details."
Example 3: Technical support - Ticket creation and escalation
Scenario
A software company wants to automate support ticket creation and escalation for critical issues.
Action: Create support ticket
Name: Create Technical Support Ticket
Description: Creates a new ticket in Zendesk when user reports a technical problem.
URL: https://company.zendesk.com/api/v2/tickets.json
Method: POST
Authentication: Basic Auth (username/token)
Content-Type: application/json
Parameters:
- customer_email (required): Customer email
- subject (required): Problem subject
- description (required): Detailed description
- priority (optional, default: normal): normal, high, urgent, low
- category (required): login, payment, bug, feature_request
- product_version (optional): Product version
Action: Urgent escalation
Name: Urgent Support Escalation
Description: Sends immediate notification to support team for critical issues that block service.
URL: https://hooks.slack.com/services/T00/B00/XXX
Method: POST
Parameters:
- ticket_id (required): Created ticket ID
- customer_info (required): Customer information
- issue_summary (required): Problem summary
- impact_level (required): Business impact level
Example 4: Marketing - Lead generation and follow-up
Scenario
A marketing agency wants to capture qualified leads and automatically insert them into CRM with personalized follow-up.
Action: Create lead in CRM
Name: Add Lead to HubSpot
Description: Creates a new contact in HubSpot when a visitor shows interest in services.
URL: https://api.hubapi.com/crm/v3/objects/contacts
Method: POST
Authentication: Bearer Token
Parameters:
- email (required): Contact email
- firstname (required): First name
- lastname (required): Last name
- company (optional): Company name
- phone (optional): Phone number
- lead_source (required): chatbot
- interest_area (required): Interest area
- budget_range (optional): Budget range
Action: Welcome email
Name: Send Lead Welcome Email
Description: Automatically sends personalized welcome email to new leads.
URL: https://api.mailgun.com/v3/domain.com/messages
Method: POST
Authentication: Basic Auth (api:API_KEY)
Parameters:
- to_email (required): Recipient
- to_name (required): Recipient name
- interest_area (required): To personalize content
- company_name (optional): For additional personalization
Example 5: Hospitality - Restaurant booking management
Scenario
A restaurant wants to automate bookings and special request management through the website assistant.
Action: Check table availability
Name: Check Restaurant Availability
Description: Checks available tables for specified date, time and number of people.
URL: https://booking.myrestaurant.com/api/availability
Method: GET
Parameters:
- date (required): Booking date (YYYY-MM-DD)
- time (required): Desired time (HH:MM)
- guests (required): Number of people
- special_requirements (optional): Special requests
Action: Create booking
Name: Book Restaurant Table
Description: Makes a confirmed booking after checking availability.
URL: https://booking.myrestaurant.com/api/reservations
Method: POST
Parameters:
- customer_name (required): Name for booking
- customer_email (required): Contact email
- customer_phone (required): Phone
- reservation_date (required): Date and time
- guests_count (required): Number of diners
- special_requests (optional): Special requests
- table_preference (optional): Table preference
Implementation considerations
Security and privacy
- Data validation: Always validate parameters before sending to APIs
- Sensitive data: Don't log personal or financial information
- Rate limiting: Respect external API limits
- Error handling: Handle errors gracefully without exposing technical details
User experience
- Immediate feedback: Always confirm when an action has been completed
- Clear information: Provide useful details (reference numbers, times, etc.)
- Error recovery: Offer alternatives when an action fails
- Important confirmations: Request confirmation for irreversible actions
Monitoring and maintenance
- Action logs: Keep track of all API calls
- Success metrics: Monitor success rates and response times
- Alerting: Configure alerts for recurring errors
- Documentation: Maintain updated documentation for each integration