Introduction

In this detailed summary, we explore the intricate process of creating an AI-powered appointment booking system.

The workflow involves multiple interconnected functions designed to capture user details, determine their time zone, fetch available time slots, and prepare the system for subsequent appointment confirmation.

This guide emphasizes clarity, structure, and technical precision, ensuring a comprehensive understanding of each component involved in the process.

The Connected Functions

The core functions of the AI agent are:

• Book Appointment

• Cancel Appointment

• Capture User Details

• Reschedule Appointment

This summary concentrates primarily on capturing user details and fetching available time slots, which are foundational to the booking process.

Step-by-Step Breakdown

1. Capturing User Details

The initial phase involves collecting essential user information:

Implementation Highlights:

• Consistency: Parameter names mirror those in the AI agent prompt for seamless integration.

• Validation: Checks if parameters already exist; if so, skips prompts.

• Memory Toggle: Utilizes conversation history to recall previous inputs, but accounts for new users or guest sessions.

2. Parameter Validation and Data Handling

The system performs conditional checks:

• If full_name is empty or contains "guest", prompt the user for their name.

• If email is missing, request the email.

• If phone_number is absent, ask for the contact number.

Once validated, each parameter is saved explicitly to ensure data integrity.

3. Triggering the Workflow

After capturing all necessary details, a function call initiates the next phase:

• Function Call: Executes the workflow to determine the user's timezone and fetch available slots.

• Workflow Components:

  • Determine User's Time Zone: Uses AI task based on phone number.

  • Set End Date: Adds a predefined number of days (e.g., 6 days) to define the window for available slots.

  • Fetch Available Time Slots: Queries based on event type ID, start, and end times.

4. Determining User's Time Zone

Methodology:

• Input: Phone number

• Output: Time zone in Region/City format (e.g., Europe/Budapest)

• Process:

  • AI task analyzes the phone number.

  • Returns matched time zone.

  • Saved into a custom user field, e.g., time_zone.

Example:

This step ensures that all subsequent time-related data aligns with the user's local time.

5. Setting the End Date for Slot Fetching

• Operation: Adds a specific number of days (e.g., 6 days) to the current date.

• Purpose: Defines the window for available appointment slots.

• Flexibility: Can be adjusted to 7, 10, or more days based on requirements.

Implementation:

Set custom field `end_time` to today + 6 days

This ensures the system fetches slots within a manageable timeframe, preventing excessive data retrieval.

6. Fetching Available Time Slots

Process Overview:

• Input Parameters:

  • Start Time: Current date/time

  • End Time: Calculated end date

  • Event Type ID: Specific to appointment category

  • User's Time Zone: For correct slot display

• Execution:

  • Calls an API or internal function to retrieve slots.

  • Uses the find available time slots by event slot method.

  • Can be customized by replacing event type ID with a bot field.

Handling Data:

• Results: Multiple time slots are returned, e.g., May 30, June 2, etc.

• Storage: All slots are saved into a data field, e.g., cal_available_time_slots.

Error Management:

• If the response exceeds 20,000 characters, reduce the number of days fetched.

• For example, fetch only 7 days instead of 11 to prevent overflow.

7. Presenting Available Slots to the User

Post-fetch Actions:

• Use an AI function result action to relay available slots.

• Message Format:

{
  "message": "The following time slots are available:",
  "slots": [list of time slots in JSON]
}

• This JSON data is then displayed to the user, enabling them to select a preferred time.

8. Next Steps and Future Enhancements

• The current workflow prepares the system for appointment confirmation.

• The next phase involves booking the appointment based on user-selected date and time.

• Future improvements may include:

  • Dynamic slot filtering

  • User notifications

  • Integration with external calendar systems

Conclusion

This comprehensive workflow exemplifies a robust, user-centric approach to appointment scheduling via AI. It meticulously captures user details, intelligently determines their timezone, fetches relevant time slots, and prepares data for seamless booking. The system's modular design allows for flexibility, scalability, and customization, ensuring it can adapt to various business needs.