Build A Personal 1:1 Conversational Bot With Microsoft Teams

Introduction

 

A conversation bot allows users to interact within multiple forms like text, adaptive cards, forms, etc., from the Microsoft Teams client.

 

A bot in Microsoft Team can be part of a one to one conversation, group chat, or a channel in a team. Each section has its own opportunities and challenges.

 

Conversational Bots 

 

Conversation bot is a series of messages between the users and the bot. There are three types of conversation within the Microsoft Team Client:

• Personal Conversation or Personal Chat: It's a one to one chat between users and bot.
• Group Conversation or Group Chat: It's between a bot and two or more users.
• Teams Channel: It's available to all channel team members, any team member can do a conversation. 

Bots behave differently depending on the type of conversation, for example:

• Personal conversation i.e. one to one conversation doesn't need @mention. All messages sent by the user directly routed to the bot.
• Teams Channel or Group Conversation requires users to use @mention to invoke the bot into the channel or group chat. 

Prerequisites

 

To build Microsoft Teams Bot, below are the necessary tools needed to install at your environment:

1. Microsoft Azure Subscription
2. Office 365 Tenant
3. Microsoft Team enable at Office 365 tenant and allowed the custom app to upload
4. Node.js (V10.* or higher)
5. NPM (V6.* or higher)
6. Gulp (V4.* or higher)
7. Yeoman (V3.* or higher)
8. Yeomen Generator of MS Teams
9. Visual Studio Code

Let's start with Azure Bot

 

To create Microsoft Team Bot requires two activities:

1. Create Microsoft Azure Bot.
2. Add a bot to MS Team-based code base. 

Step 1 - Create Microsoft Azure Bot

1. Browse the https://portal.azure.com with your work or school account.
2. Click "+Create Resource"
3. Select "AI + Machine Learning"
4. Select "Web App Bot"

Step 2 - Complete all required information and click create to provision the bot

1. Bot handle:- its unique name of your bot.
2. Subscription:- Select a valid available subscription.
3. Resource Group:- select an appropriate resource group or create a new one as per requirement. 
4. Location:- Select your closed azure location or the preferred location.
5. Pricing:- Select the required pricing tier or F0 as free for POC activity.
6. App Name:-  Leave it as-is for this demo purpose.
7. Service Plan:- Leave it as-is for this demo purpose.
8. Bot Template:- Leave it as-is for this demo purpose.
9. Application insight: Select off for this demo purpose
10. Microsoft App Id and Password:- Leave it as-is for this demo purpose.
11. Click create to provision. 

 

 

Step 3 - Enable Microsoft Team Channel for your Bot

 

In order for the bot to interact with MS Teams, you must enable the Teams Channel.

 

 

To complete the process, MS Teams and Webchat should be available and should be listed in your channel lists.

 

 

Step 4 - Create a Microsoft Teams App

 

In this section, we are going to create a Node.js project

• Open the Command Prompt and navigate to the desired directory to create a project.
• Run the Yeomen Generator for Microsoft Teams
  
  yo teams  

• Yeomen generator will ask the following question:

 

Step 5 - Update Default Bot Code

 

Here we will find a 1:1 conversation response. 

 

Browse Interactive ChatBotBot file  

 

 

Step 6

 

To implement this functionality, locate and open the ./src/app/interactiveChatBotBot/InteractiveChatBotBot.ts file and add the following method to the InteractiveChatBotBot class:

 

Add the below code after the import statement 

import * as Util from "util";
const TextEncoder = Util.TextEncoder;

 Add Message Factory object reference to the existing botbuilder package:

import {
  StatePropertyAccessor,
  CardFactory,
  TurnContext,
  MemoryStorage,
  ConversationState,
  ActivityTypes,
  TeamsActivityHandler,
  MessageFactory,
} from 'botbuilder';

Locate the handler onMessage() within the constructor().

 

Locate and replace the line if (text.startsWith("hello")) { in the onMessage() handler with the following code:

if (text.startsWith("onetoonepersonalchat"))
  await this.handleMessageMentionMeOneOnOne(context);
  return;
} else if (text.startsWith("hello")) {

Add the following method to the class:

private async handleMessageMentionMeOneOnOne(context: TurnContext): Promise<void> {
  const mention = {
    mentioned: context.activity.from,
    text: `<at>${new TextEncoder().encode(context.activity.from.name)}</at>`,
    type: "mention"
  };

  const replyActivity = MessageFactory.text(`Hi ${mention.text} from a one to one personal chat.`);
  replyActivity.entities = [mention];
  await context.sendActivity(replyActivity);
}

Step 7 - Update Project Environment Variable

 

Locate and open the file ./.env.

 

Locate the following section in the file and set the values of the two properties that you obtained when registering the bot:

# App Id and App Password fir the Bot Framework bot
MICROSOFT_APP_ID=
MICROSOFT_APP_PASSWORD=

Step 8 - Update Manifest File

• Browse manifest.json file  ./src/maifest/manisfest.json
• Update below properties:-
• ID: Replace with Azure bot ID
• Version: 1.0.0
• PackageName   : InteractiveBotBot
• Bot Property

"bots": [
    {
      "botId": "b0edaf1f-0ded-4744-ba2c-113e50376be6",
      "needsChannelSelector": true,
      "isNotificationOnly": false,
      "scopes": [
        "team",
        "personal"
      ],
      "commandLists": [
        {
          "scopes": [
            "team",
            "personal"
          ],
          "commands": [
            {
              "title": "Help",
              "description": "Shows help information"
            },
            {
              "title": "mentionme",
              "description": "Sends message with @mention of the sender"

            }
          ]
        }
      ]
    }
  ],

Step 9 - Test and Run the Bot

 

From VSCode select View -> Terminal and Hit below command

gulp ngrok-serve

Ngrok-serve builds your project and starts running at a local webserver. Ngrok start with random subdomain and create  secure URL for the local webserver

 

Microsoft Team requires all content to be displayed with an HTTPS request. Local debugging requires local HTTP webserver. Ngrok creates a secure routable URL to HTTP webserver to debug the HTTPS application securely.

 

 

Ngrok created a temporary URL "14dceed6815b.ngrok.io". The same needs to be updated at Azure Bot with Message endpoint as per the below screenshot.

• Select Setting under Bot Management
• Update Messaging End Point with Ngrok URL
• Click save to update.

Step 9 - Upload Manifest File

 

Navigate to manifest folder /src/manifest. Select both .png files and manifest.jon file and zip all three files. Give the name of zip file "manifest"

 

Login to https://teams.microsoft.com 

1. Navigate to App Studio
2. Select Import an existing app
3. Select the manifest.zip file
4. Bot name with the icon will appear at the screen

 

Step 10 - Install Custom App to Teams Solution

 

Using the app bar navigation menu, select the Mode added apps button. Then select Browse all apps followed by Upload for me or my teams.

 

 

In the file dialog that appears, select the Microsoft Teams package in your project. This app package is a ZIP file that can be found in the project's ./package folder.

 

Click Add or Select app to navigate to chat with the bot 

 

 

Select the MentionMe command, or manually type mentionme in the compose box, then press enter.

 

After a few seconds, you should see the bot respond mentioning the user you are signed in with,

 

 

Code base exists at this location.

 

 I hope you have enjoyed and learned something new in this article. Thanks for reading and stay tuned for the next article.