Calendar Secretary Agent - Complete Tutorial

In this tutorial, you'll learn to create a complete Secretary Agent for a dental office, capable of:

Querying availability in Google Calendar
Scheduling appointments directly via WhatsApp
Performing post-service analysis
Automatically recording interactions in Google Sheets

Prerequisites

Configured SipPulse AI account
Configured WhatsApp Business (see guide)
Google account with active Google Calendar
Make.com account
Google Sheets for data recording

1. Setting Up Automations in Make

1.1 Webhook to Query Availability

First, let's create a webhook that queries available times in the calendar.

Access Make and create a new scenario
Add a Webhook module as trigger:
- Click Create a webhook
- Name it query_dentist_calendar
- Copy the generated URL (will be used in the agent)

Configure Google Calendar:

Add module Google Calendar → Search Events
Configure Google credentials

Important parameters:

Calendar ID: primary (or specific office ID)
Time Min: {{query.start_date}} (comes from webhook)
Time Max: {{query.end_date}} (comes from webhook)
Max Results: 50

Processing and Response:
- Add module Tools → Set Variables to process data
- Add module Webhooks → Webhook Response to return JSON:

json

{
  "busy_times": [
    {
      "start": "2024-01-15T09:00:00",
      "end": "2024-01-15T10:00:00",
      "title": "Appointment - John Silva"
    }
  ],
  "available_times": [
    "09:00", "10:00", "11:00", "14:00", "15:00", "16:00"
  ],
  "queried_date": "2024-01-15"
}

1.2 Webhook to Schedule Appointment

Create another scenario in Make
Add Webhook trigger:
- Name: schedule_dentist_appointment
- Copy the URL

Configure Google Calendar Create Event:

Module: Google Calendar → Create an Event

Parameters:

Calendar ID: primary
Summary: Appointment - {{query.patient_name}}
Start Date: {{query.start_date}}
End Date: {{query.end_date}}
Description: Scheduled via WhatsApp
Location: Dr. Silva's Office

Confirmation response:

json

{
  "success": true,
  "event_id": "{{google_calendar_event_id}}",
  "message": "Appointment successfully scheduled for {{query.patient_name}}"
}

2. Creating the Secretary Agent

2.1 Basic Configuration

Access Agents on the SipPulse AI platform
Click "New Agent"
Configure profile:
- Name: Dr. Silva's Secretary
- Description: Virtual assistant for dental appointment scheduling
- Model: GPT-4.1 (recommended for complex tasks)

2.2 System Instructions

markdown

# IDENTITY
You are Marina, the virtual secretary for Dr. Silva's dental office. You are professional, caring, and expert in scheduling.

# PERSONALITY
- Always polite and empathetic
- Efficient in scheduling
- Well-versed in office procedures
- Patient with client questions
- Proactive in offering solutions

# OFFICE HOURS
- Monday to Friday: 8am to 6pm
- Saturday: 8am to 12pm
- Sunday: Closed
- Appointment duration: 30 minutes each
- Lunch break: 12pm to 2pm

# AVAILABLE PROCEDURES
- Routine consultation (30 min)
- Cleaning (45 min)
- Root canal treatment (60 min)
- Extraction (45 min)
- Whitening (60 min)

# SCHEDULING FLOW
1. Greet the patient and ask for their name
2. Identify the desired type of procedure
3. Query availability using the query_calendar tool
4. Present available time options
5. Confirm details with the patient
6. Schedule the appointment using the schedule_appointment tool
7. Provide important information (address, recommendations)

# IMPORTANT RULES
- Always confirm full name and phone before scheduling
- Don't schedule outside office hours
- For long procedures (>45min), preferably in the morning
- Always ask if it's a first visit to the office
- Offer rescheduling if no time available on requested day

# OFFICE INFORMATION
- Address: Flower Street, 123 - Downtown
- Phone: (11) 99999-9999
- Dr. Silva - CRO 12345
- Free parking available
- Accepts card and PIX

# FINAL INSTRUCTIONS
- For dental emergencies, advise calling directly
- Always end by asking if there are any other questions
- Be clear about cancellation policy (24h advance notice)

2.3 Enabling Tools

Tool 1: Query Calendar

Click "+ New tool"
Select "API Integration"
Configure:
- Name: query_calendar
- Description: Queries available times in the office's Google Calendar
- Structure:
  json
```
{
  "start_date": "string",
  "end_date": "string"
}
```
- URL: Paste the Make webhook URL (query)
- Method: GET
- Mode: Query Parameters

Tool 2: Schedule Appointment

Add another tool
Configure:
- Name: schedule_appointment
- Description: Schedules a new appointment in Google Calendar
- Structure:
  json
```
{
  "patient_name": "string",
  "phone": "string",
  "procedure": "string",
  "start_date": "string",
  "end_date": "string"
}
```
- URL: Paste the Make webhook URL (schedule)
- Method: POST
- Mode: Body

3. Creating Post-Service Analysis

3.1 New Structured Analysis

Access "Structured Analyses"
Click "New Analysis"
Configure:
- Name: dental_secretary_service
- Description: Evaluates service quality and scheduling success
- Model: GPT-4.1-mini

3.2 Analysis Structure

Configure the following fields:

Name: dental_secretary_service

Description: Post-service analysis to evaluate appointments and patient satisfaction at the dental office.

Structure:

appointment_scheduled
- Type: Boolean
- Name: appointment_scheduled
- Description: Whether an appointment was effectively scheduled
patient_name
- Type: Text
- Name: patient_name
- Description: Full name of the patient served
procedure_type
- Type: Text
- Name: procedure_type
- Description: Type of procedure requested (Routine consultation, Cleaning, Root canal treatment, Extraction, Whitening, Not defined)
scheduled_date
- Type: Text
- Name: scheduled_date
- Description: Scheduled date and time (format YYYY-MM-DD HH:mm)
service_satisfaction
- Type: Text
- Name: service_satisfaction
- Description: Satisfaction assessment based on conversation tone (Very satisfied, Satisfied, Neutral, Dissatisfied, Very dissatisfied)
no_scheduling_reason
- Type: Text
- Name: no_scheduling_reason
- Description: If no appointment was made, what was the reason
first_visit
- Type: Boolean
- Name: first_visit
- Description: Whether this is the patient's first time at the office
notes
- Type: Text
- Name: notes
- Description: Important observations about the service

3.3 Activating Post-Analysis in Agent

Edit the created agent
Go to "Post-Analysis" tab
Enable the toggle for the dental_secretary_service analysis
Save the agent

4. Connecting to WhatsApp

4.1 Channel Configuration

In the agent, click "⋮ → Deploy"
On the WhatsApp card, click "Connect"
Follow the authentication process (see detailed guide)

4.2 Configure Preferences

Click "⋮ → Preferences" on the WhatsApp card
Configure:
- Inactivity closure: 10 minutes
- Closing message: "Thank you for reaching out! If you need anything else, I'll be here. Have a great day! 😊"
- Audio support: Enabled with model pulse-precision-v3

5. Configuring Webhook for Google Sheets

5.1 Creating Webhook on the Platform

Access "Webhooks" in the top menu
Click "Create"
Configure:
- URL: Make webhook URL (which we'll create next)
- Secret: my_secret_key_dentist_2024
- Events: Check only thread.closed

5.2 Make Automation for Google Sheets

Create new scenario in Make
Add Webhook module as trigger
Configure Google Sheets:
- Module: Google Sheets → Add a Row
- Select your spreadsheet
- Configure columns:

Date/Time: {{webhook.data.timestamp}}
Patient Name: {{webhook.data.post_analysis.patient_name}}
Appointment Scheduled: {{webhook.data.post_analysis.appointment_scheduled}}
Procedure: {{webhook.data.post_analysis.procedure_type}}
Scheduled Date: {{webhook.data.post_analysis.scheduled_date}}
Satisfaction: {{webhook.data.post_analysis.service_satisfaction}}
First Visit: {{webhook.data.post_analysis.first_visit}}
Notes: {{webhook.data.post_analysis.notes}}

5.3 Security Validation in Make

Add a Tools → Set Variables module before Google Sheets to validate the webhook:

javascript

// HMAC Validation (adapted for Make)
const crypto = require('crypto');
const secret = 'my_secret_key_dentist_2024';
const signature = headers['x-signature'];
const timestamp = headers['x-timestamp'];
const payload = JSON.stringify(webhook_data);

const expectedSignature = crypto
  .createHmac('sha256', secret)
  .update(`${payload}:${timestamp}`)
  .digest('hex');

if (expectedSignature !== signature) {
  throw new Error('Invalid signature');
}

6. Testing the Complete System

6.1 Testing on WhatsApp

Send a message to the connected number:

User: Hi, I'd like to schedule an appointment
Marina: Hello! I'm Marina, Dr. Silva's virtual secretary. It's my pleasure to help you! 😊

May I have your full name, please?

User: Maria Santos
Marina: Nice to meet you, Maria! What type of procedure would you like to schedule?

- Routine consultation (30 min)
- Cleaning (45 min)
- Root canal treatment (60 min)
- Extraction (45 min)
- Whitening (60 min)

User: Routine consultation
Marina: Perfect! When would you like to schedule it? Can you tell me your preferred date?

6.2 Verify Integrations

Calendar: Check if events appear in Google Calendar
Sheets: Confirm data is recorded after session closure
Analysis: View results in the thread interface

7. Monitoring and Improvements

7.1 Important Metrics

Track in Google Sheets:

Conversion rate (appointments/services)
Most requested procedure types
Peak demand times
Patient satisfaction

7.2 Suggested Optimizations

Add reminders: Create automation to send reminders 1 day before
Cancellations: Implement tool to cancel/reschedule
Waiting list: System to notify about newly available slots
Proactive templates: Post-appointment follow-up messages

Conclusion

Now you have a complete WhatsApp scheduling system that:

✅ Queries availability in real-time ✅ Automatically schedules in Google Calendar ✅ Analyzes service quality ✅ Records data for later analysis ✅ Maintains complete interaction history

The system is ready to receive patients and can be easily adapted for other types of businesses that rely on scheduling

Agents

Configuration

Tools

Advanced

Deploying Agents

Settings

Calendar Secretary Agent - Complete Tutorial

Prerequisites

1. Setting Up Automations in Make

1.1 Webhook to Query Availability

1.2 Webhook to Schedule Appointment

2. Creating the Secretary Agent

2.1 Basic Configuration

2.2 System Instructions

2.3 Enabling Tools

Tool 1: Query Calendar

Tool 2: Schedule Appointment

3. Creating Post-Service Analysis

3.1 New Structured Analysis

3.2 Analysis Structure

3.3 Activating Post-Analysis in Agent

4. Connecting to WhatsApp

4.1 Channel Configuration

4.2 Configure Preferences

5. Configuring Webhook for Google Sheets

5.1 Creating Webhook on the Platform

5.2 Make Automation for Google Sheets

5.3 Security Validation in Make

6. Testing the Complete System

6.1 Testing on WhatsApp

6.2 Verify Integrations

7. Monitoring and Improvements

7.1 Important Metrics

7.2 Suggested Optimizations

Conclusion

Configuration

Tools

Advanced

Deploying Agents

Calendar Secretary Agent - Complete Tutorial ​

Prerequisites ​

1. Setting Up Automations in Make ​

1.1 Webhook to Query Availability ​

1.2 Webhook to Schedule Appointment ​

2. Creating the Secretary Agent ​

2.1 Basic Configuration ​

2.2 System Instructions ​

2.3 Enabling Tools ​

Tool 1: Query Calendar ​

Tool 2: Schedule Appointment ​

3. Creating Post-Service Analysis ​

3.1 New Structured Analysis ​

3.2 Analysis Structure ​

3.3 Activating Post-Analysis in Agent ​

4. Connecting to WhatsApp ​

4.1 Channel Configuration ​

4.2 Configure Preferences ​

5. Configuring Webhook for Google Sheets ​

5.1 Creating Webhook on the Platform ​

5.2 Make Automation for Google Sheets ​

5.3 Security Validation in Make ​

6. Testing the Complete System ​

6.1 Testing on WhatsApp ​

6.2 Verify Integrations ​

7. Monitoring and Improvements ​

7.1 Important Metrics ​

7.2 Suggested Optimizations ​

Conclusion ​

Calendar Secretary Agent - Complete Tutorial

Prerequisites

1. Setting Up Automations in Make

1.1 Webhook to Query Availability

1.2 Webhook to Schedule Appointment

2. Creating the Secretary Agent

2.1 Basic Configuration

2.2 System Instructions

2.3 Enabling Tools

Tool 1: Query Calendar

Tool 2: Schedule Appointment

3. Creating Post-Service Analysis

3.1 New Structured Analysis

3.2 Analysis Structure

3.3 Activating Post-Analysis in Agent

4. Connecting to WhatsApp

4.1 Channel Configuration

4.2 Configure Preferences

5. Configuring Webhook for Google Sheets

5.1 Creating Webhook on the Platform

5.2 Make Automation for Google Sheets

5.3 Security Validation in Make

6. Testing the Complete System

6.1 Testing on WhatsApp

6.2 Verify Integrations

7. Monitoring and Improvements

7.1 Important Metrics

7.2 Suggested Optimizations

Conclusion