Building a Chatbot with Azure Bot Service

Posted on by Panther

Introduction

Chatbots have gained massive popularity in recent times, thanks to their ability to automate customer interactions and improve user engagement. With a chatbot, businesses can easily handle customer queries, provide personalized recommendations and offer round-the-clock customer support. However, building a chatbot from scratch can be a daunting task, especially for developers with no prior experience. This is where Azure Bot Service comes in. Azure Bot Service is a cloud-based platform that simplifies the process of building chatbots, enabling developers to focus on designing conversational experiences.

In this tutorial, we will show you how to build a chatbot with Azure Bot Service. We will cover various topics such as creating the bot’s backend, integrating the bot with Facebook Messenger, and publishing the bot to Azure. Let’s get started.

Prerequisites

Before we begin, ensure that you have the following installed:

  • An active Azure subscription
  • Node.js installed on your machine
  • Visual Studio Code (or any code editor of your choice) installed
  • The Bot Framework Emulator installed

Step 1: Create a new Bot Service

The first step in building a chatbot with Azure Bot Service is creating a new bot service. To create a new bot service, follow these steps:

  1. Log in to the Azure portal.
  2. Click on “Create a resource” and search for “Bot Service”.
  3. From the search results, select “Bot Service” and click “Create”.
  4. Fill in the necessary details such as subscription, resource group, and bot name.
  5. Choose a runtime stack (Node.js for our case) and template (basic bot).
  6. Click “Create” to create the bot service.

Step 2: Create the bot’s backend

After creating the bot service, the next step is to create the bot’s backend. The bot’s backend will handle all the logic for processing user queries and generating responses. In our case, we will be using Node.js as the programming language for the backend. To create the backend, follow these steps:

  1. Open Visual Studio Code (or any code editor of your choice) and create a new folder for the bot.
  2. Open a new terminal in Visual Studio Code and navigate to the bot folder.
  3. Run the following command to create a new Node.js project: 
    npm init
  4. Once the project is created, install the following dependencies: 
    npm install botbuilder botbuilder-ai --save

    The botbuilder package provides the core functionality for developing chatbots, while the botbuilder-ai package provides support for integrating natural language processing services such as LUIS.

  5. Next, create a new file named index.js in the bot folder and add the following code:

    const { ActivityTypes, MessageFactory } = require('botbuilder');
const { LuisRecognizer } = require('botbuilder-ai');

const LuisAppId = '<LUIS_APP_ID>';
const LuisAPIKey = '<LUIS_API_KEY>';
const LuisAPIHostName = '<LUIS_API_HOSTNAME>';

const luisConfig = {
   applicationId: LuisAppId,
   endpointKey: LuisAPIKey,
   endpoint: `https://${LuisAPIHostName}`
};

const luisRecognizer = new LuisRecognizer(luisConfig);

const { ActivityHandler } = require('botbuilder');

class EchoBot extends ActivityHandler {
   constructor() {
       super();

       this.onMessage(async (context, next) => {
           const { text } = context.activity;

           const results = await luisRecognizer.recognize(text);

           const topIntent = results.luisResult.topScoringIntent.intent;

           let response = '';
           switch (topIntent) {
               case 'greeting':
                   response = 'Hi, how can I help you today?';
                   break;
               case 'goodbye':
                   response = 'Goodbye, it was nice talking to you!';
                   break;
               default:
                   response = 'Sorry, I did not understand your request.';
                   break;
           }

           const message = MessageFactory.text(response);
           await context.sendActivity(message);

           await next();
       });
   }
}

module.exports.EchoBot = EchoBot;

    In this code, we first import the necessary packages and create an instance of the LuisRecognizer class, which will be used for natural language processing. We then create a new class called EchoBot that extends the ActivityHandler class, which provides built-in methods for handling user interactions such as onMessage. In the onMessage method, we use the LuisRecognizer instance to recognize the user’s intent and generate an appropriate response.

  6. Next, create a new file named index.bot in the bot folder and add the following code:

    {
   "type": "abs",
   "id": "echo-bot",
   "name": "Echo Bot",
   "version": "1.0",
   "services": [
       {
           "type": "endpoint",
           "name": "development",
           "endpoint": "http://localhost:3978/api/messages",
           "appId": "",
           "appPassword": "",
           "id": "http://localhost:3978/api/messages"
       }
   ]
}

    In this code, we define the bot’s metadata such as the name and version. We also define an endpoint service that points to the URL of the bot’s backend.

Step 3: Test the bot locally

After creating the bot’s backend, the next step is to test the bot locally. To test the bot locally, follow these steps:

  1. Open a new terminal in Visual Studio Code and navigate to the bot folder.
  2. Run the following command to start the bot: 
    node index.js
  3. Once the bot is running, open the Bot Framework Emulator and create a new bot configuration.
  4. Enter the URL of the bot’s backend (e.g. `http://localhost:3978/api/messages`) and click “Save and Connect”.
  5. Send a message to the bot and verify that it responds correctly.

Step 4: Integrate the bot with Facebook Messenger

After testing the bot locally, the next step is to integrate the bot with Facebook Messenger. To integrate the bot with Facebook Messenger, follow these steps:

  1. Log in to the Azure portal and navigate to the bot service.
  2. Click on “Channels” and select “Facebook Messenger”.
  3. Follow the on-screen instructions to configure the Facebook Messenger integration. You will need to provide your Facebook credentials and create a new Facebook page for the bot if you haven’t already.
  4. Once the integration is configured, copy the Page Access Token from the Facebook Messenger configuration page.
  5. Open the .env file in the bot folder and add the following line: 
    MicrosoftAppPassword=<PAGE_ACCESS_TOKEN>
  6. Save the .env file.

Step 5: Publish the bot to Azure

After integrating the bot with Facebook Messenger, the final step is to publish the bot to Azure. To publish the bot to Azure, follow these steps:

  1. Log in to the Azure portal and navigate to the bot service.
  2. Click on “Build” and select “Deploy”.
  3. Follow the on-screen instructions to deploy the bot to Azure.

Conclusion

In this tutorial, we showed you how to build a chatbot with Azure Bot Service. We covered various topics such as creating the bot’s backend, integrating the bot with Facebook Messenger, and publishing the bot to Azure. With Azure Bot Service, building a chatbot has never been easier. With the knowledge gained in this tutorial, you can now create your own chatbot and automate your customer interactions.

Related Post