Booky - The Multi Agentic Appointment Booking System

Booky is an intelligent, multi-agent AI-powered appointment booking system designed to streamline scheduling through a conversational interface. It leverages advanced AI models to understand user intent, check availability, book appointments, update records, and send confirmation emails—automating the entire appointment management process.

---

Features

• AI-driven conversational booking powered by Google Gemini model
• Real-time availability checking of appointment slots in a SQLite database
• Automated booking that updates the database and Google Sheets
• Email confirmation sent to users upon successful booking
• Web interface built with Flask for user interaction
• Modular, multi-agent architecture for clear separation of concerns

---

Architecture Overview

START
  |
  |-- Load Configuration & .env
  |
  |-- Initialize Flask App
  |
  |-- Initialize Database
  |        |
  |        '--- Seed time slots if empty
  |
  |-- Initialize Google Sheets API
  |
  |-- User Opens Website (Serves index.html)
  |
  |-- User Sends Chat Message
           |
           '-- Gemini Model Processes Message
                    |
                    '-- Check Intent: [Availability] or [Booking]
                           |                      |
             [get_available_slots(date)]    [book_appointment(slot_time, name, email)]
                           |                      |
              Return available slots    Update DB, Log to Sheets, Send Email
                           |                      |
                 User picks a slot? <------------+
                           |
                   Proceed to booking
                           |
                     Booking confirmation
                           |
                      Send final reply
                           |
                          END

    

---

Getting Started

Prerequisites

• Python 3.8 or higher
• SQLite (usually included with Python)
• Google Cloud project with Google Sheets API enabled
• Gmail account (or another SMTP-enabled email) for sending confirmation emails
• .env file with API keys and credentials

---

Installation and Setup

1. Clone the repository

   git clone https://github.com/Vatsa10/Booky.git
   cd Booky
   

2. Create and activate a Python virtual environment

   python -m venv venv
   source venv/bin/activate      # For Linux/macOS
   venv\Scripts\activate         # For Windows
   

3. Install dependencies

   pip install -r req.txt
   

4. Google Cloud and API configuration

   • Go to Google Cloud Console
   • Create/select a project
   • Enable Google Sheets API and Gmail API
   • Create a Service Account with Editor permission
   • Download the JSON key file for the Service Account

5. Set up Google Sheets

   • Create a new spreadsheet
   • Share it with the service account’s email address with Editor permission
   • Copy the spreadsheet ID from the Google Sheets URL

6. Create a .env file in the root directory with the following

   GEMINI_API_KEY=your_gemini_api_key_here
   GOOGLE_SHEETS_CREDENTIALS=path/to/service-account-credentials.json
   GOOGLE_SPREADSHEET_ID=your_spreadsheet_id_here
   EMAIL_ADDRESS=your_email@example.com
   EMAIL_PASSWORD=your_email_password_or_app_password
   BUSINESS_NAME=YourBusinessName
   

---

Running the Application

Run the Flask web server and initialize the database seed:

python app.py

This will start the server on:

http://localhost:5000

---

Usage

• Access the URL in your browser
• Type messages in the chat interface to:
  • Check available appointment slots
  • Book an appointment by providing name and email
• Receive booking confirmation emails automatically

---

Project Structure

booky/
│
├── app.py                  # Main Flask app with AI agent logic
├── appointments.db         # SQLite database file (auto-created)
├── public/
│   ├── index.html          # Frontend UI for chat
│   └── ...                 # Static assets (CSS, JS)
├── credentials.json        # Google API service account credentials
├── requirements.txt        # Python dependencies
└── .env                    # Environment variables, not included in repo

---

Enjoy effortless AI-powered appointment scheduling with Booky!