はじめに
対話型AIプラットフォーム
チャットボットの概要
自然言語処理(NLP)
ボットの概念と用語
クイックスタートガイド
プラットフォームへのアクセス
ボットビルダーの操作
リリースノート
最新バージョン(英語)
以前のバージョン(英語)
廃止機能(英語)
コンセプト
設計
ストーリーボード
ダイアログタスク
ダイアログタスクとは
ダイアログビルダー
ノードタイプ
インテントノード
ダイアログノード
エンティティノード
フォームノード
確認ノード
ロジックノード
ボットアクションノード
サービスノード
Webhookノード
スクリプトノード
グループノード
エージェント転送ノード
ユーザープロンプト
音声通話プロパティ
イベント ハンドラー
ナレッジグラフ
ナレッジグラフの抽出
ナレッジグラフの構築
ボットにナレッジグラフを追加
グラフの作成
ナレッジグラフの構築
既存のソースからFAQを構築
特性、同義語、停止用語
変数ネームスペースの管理
更新
ノード間の質問と回答の移動
用語の編集と削除
質問と応答の編集
ナレッジグラフの分析
通知タスク
スモールトーク
デジタルスキル
デジタルフォーム
デジタルビュー
デジタルビューとは
パネル
ウィジェット
トレーニング
トレーニングとは
機械学習
機械学習とは
モデル検証
ファンダメンタルミーニング
ナレッジグラフ
示唆
ランキングおよび解決
NLPの詳細設定
NLPのガイドライン
インテリジェンス
インテリジェンスとは
コンテキスト
コンテキストインテント
割り込み
複数インテントの検出
エンティティの変更
デフォルトの会話
センチメント管理
トーン分析
テストとデバッグ
ボットと会話
発話テスト
バッチテスト
会話テスト
デプロイ
チャネル
公開
分析
ボットの分析
NLPメトリクス
会話フロー
Usage Metrics
封じ込め測定
カスタムダッシュボード
カスタムダッシュボードとは
メタタグ
カスタムダッシュボードとウィジェット
ユニバーサルボット
ユニバーサルボットとは
ユニバーサルボットの定義
ユニバーサルボットの作成
ユニバーサルボットのトレーニング
ユニバーサルボットのカスタマイズ
他言語の有効化
ストア
プラントと使用
Overview
Usage Plans
Support Plans
Invoices
管理
ボット認証
複数言語対応ボット
個人を特定できる情報の編集
ボット変数の使用
IVRのシステム連携
一般設定
ボット管理
ハウツー
会話スキルの設計
バンキングボットを作成
バンキングボット – 資金の振り替え
バンキングボット – 残高を更新
ナレッジグラフを構築
スマートアラートの予約方法
デジタルスキルの設計
デジタルフォームの設定方法
デジタルビューの設定方法
データテーブルのデータの追加方法
データテーブルのデータの更新方法
Add Data from Digital Forms
ボットのトレーニング
示唆の使用方法
インテントとエンティティのパターンの使用方法
コンテキスト切り替えの管理方法
ボットのデプロイ
エージェント転送の設定方法
ボット関数の使用方法
コンテンツ変数の使用方法
グローバル変数の使用方法
Web SDK Tutorial(英語)
Widget SDK Tutorial(英語)
ボットの分析
カスタムダッシュボードの作成方法
カスタムタグを使ってフィルタリング
管理
ボット管理者コンソール
ダッシュボード
ユーザーの管理
ユーザーの管理
グループの管理
ロール管理
ボット管理モジュール
登録
ユーザーの招待
招待状の一括送信
ユーザーデータのインポート
Active Directoryからユーザーを同期
セキュリティ/コンプライアンス
シングル サインオンの使用
セキュリティ設定
Billing(日本未対応)
  1. ホーム
  2. Docs
  3. Virtual Assistants
  4. SDKs
  5. BotKit SDK Tutorial – Agent Transfer

BotKit SDK Tutorial – Agent Transfer

This tutorial is an end-to-end walkthrough for setting up and running an instance of the Kore.ai BotKit SDK which is a collection of libraries that you can use for more control of the messages and webhook events for Kore.ai assistants for your web service. For this tutorial, you will use the sample assistant called Airlines Airlines Travel Planning that you can install as a new assistant, import a dialog task for agent transfer, generate authentication tokens, test the assistant using ngrok as a test callback server, and access the agent transfer to a live agent using LiveChat.

Note: This tutorial may require a newer version of BotSDK files. If the downloaded BotKit SDK files do not contain LiveChat.js and LiveChatAPI.js files, contact support.

Prerequisites

To complete this tutorial, you will have to set up and configure the following prerequisites:

  • LiveChat account – Subscribe to or set up a free 30-day trial. You will need the license number from the LiveChat account to configure the BotKit SDK settings for this tutorial.
  • Install and run ngrok as your test callback server. The test callback server application, ngrok, is used to simulate your callback server application.
  • Install node.js on your computer – The BotKit SDK requires node.js (version 10 or above) to communicate between your server and the Kore.ai XO Platform.

To get your LiveChat license code for your BotKit SDK configuration: 

  1. Log into your LiveChat account and click the Settings icon on the left navigation bar. 
  2. You will be taken to the Website Channel Settings page and you will find the License ID right underneath the area that lets you Install LiveChat Code Manually, as illustrated below.

Note: Make sure url-template and node-schedule modules (and any other modules as referenced in js file related to LiveChat) are installed on your computer.

 

Installing and running ngrok

To download the test callback server application, click https://ngrok.com/download, and then download and install the ngrok file for your operating system.

In a Terminal Window, start ngrok to monitor port 8003 using the following command:

ngrok http 8003

Copy the https Forwarding URL as shown in the previous illustration, for example, https://1b47f1d2.ngrok.io. Leave the terminal window open and ngrok running.

Installing node.js

The BotKit SDK requires node.js (version 10 or above) to run on the same server where the SDK is installed. Go to https://nodejs.org/en/download/, and then select your OS as a .pkg file type for Mac and .msi file type for Windows.

In a Terminal window, run the node -v command to verify installation and version, for example, v6.10.2 as shown in the following illustration.

Installing and Configuring Your Assistant

With the prerequisites installed and running, you are ready to install the Airlines Travel Planning assistant and then import the Agent Transfer dialog task into it. In this section, you will:

  1. Install the Travel Planning assistant 
  2. Create a Dialog Task with an Agent Transfer node
  3. Create an App Registration
  4. Generate Security Credentials
  5. Install and Configure the BotKit SDK – config.json, LiveChat.js, and LiveChatAPI.js files

Installing the Airlines Travel Planning Assistant

Kore.ai provides ready to use sample assistants that you can install into your instance of XO Platform. For this tutorial, you will need to install the Airlines Travel Planning assistant.

To install the Airlines Travel Planning assistant

  1. Log on to XO Platform, and then click New Bot > Start from the Store.
  2. Point to Airlines Airlines Travel Planning, and then click Install. The Airlines Airlines Travel Planning_KORE NAME_Instance# The installed successful message is displayed and the sample assistant is added to the left-hand navigation menu.

Once the sample assistant is installed, create an Agent Transfer dialog task. Learn more about the Dialog Builder and with the Agent Transfer node.

Registering Your App and Generating Client Credentials

To register your BotKit SDK with the Kore.ai XO Platform, you will need to create an app registration, and then generate security credentials used to securely communicate between the SDK installed on your server platform and the Kore.ai XO Platform.

Creating an App Registration

  1. In the XO Platform, go to Build > Integrations > BotKit
  2. Next to the App name drop-down list, click Add.
  3. Enter a name for the app and then click Next.
  4. The Client ID and Client Secret details appear. Click Done.
  5. In the Callback URL field, enter the URL for your application that you copied from the Forwarding URL from the ngrok application that you installed and ran as part of the prerequisites, for example,  https://1bXXXXd2.ngrok.io.
    Note: Each running instance of ngrok will generate a session unique URL. You need to update the Callback URL field if you stop and restart ngrok.
  6. In the Events section, select:
    1. OnMessage – An event is triggered when a message from a user is received.
    2. OnAgentTransferNode – An event is triggered when an Agent Transfer node is encountered in the dialog task flow.
  7. Click Create Credentials. The new app registration along with client credentials are displayed in the SDK Configuration section.
  8. Click Save. The Successfully subscribed message is displayed.

To configure the BotKit SDK settings, you will need to:

  • Copy the Client ID, for example, cs-6e5eb411-XXXX-XXXX-9ba2-ccc748d4f719.
  • View, and then copy the Client Secret, for example, dTTmuTOgs8SOXXXXXXXXouE0EPy28fXojOJWedOEF6s=.

Publishing the Assistant

A developer can use any installed assistant immediately for testing as it is available to the owner on save. If you only want to run and test this assistant for this tutorial, you can skip to the next section, Installing and Configuring the BotKit SDK.

To enable the assistant for other users, you must publish it so the Admin can approve and deploy the assistant for other users. For more information, see Publishing an Assistant.

After you publish an assistant in XO Platform, an Admin must complete one of the following deployments tasks in the Admin Console depending on whether your assistant’s Target Audience is set to Enterprise Users or the General Public.

  • For Enterprise Users:
    1. In the Admin Console, in the Bots Management module, on the Enterprise Bots page, click the Ellipses icon for the Agent Transfer assistant that you want to deploy, and then click Manage bot tasks. The Manage Bot Tasks dialog is displayed.
    2. Expand the Bot Tasks field to view the available and deployed tasks.
    3. Select all tasks of the assistant and then click Confirm.
    4. In the Manage bots tasks dialog, click Confirm. The Bot status changed successfully message is displayed.
    5. On the Enterprise Bots page, click the Ellipses icon for the Agent Transfer bot that you want to assign enterprise users to, and then click Bot & task assignments. The Bot & task assignments dialog is displayed.
    6. Assign the assistant to the selected users in the enterprise.

  • For General Public Users:
    1. In the Admin Console, in the Bots Management module, on the Consumer Bots page, click the Ellipses icon for the Agent Transfer assistant that you want to deploy, and then click Manage bot tasks. The Manage Bot Tasks dialog is displayed.
    2. Expand the Bot Tasks field to view the available and deployed tasks.
    3. Select all tasks of the bot and then click Confirm. The Bot status changed successfully message is displayed.

Installing and Configuring the BotKit SDK

You need to download and then uncompress the Kore.ai BotKit SDK from https://github.com/Koredotcom/BotKit. Be sure to download the latest version that contains the LiveChat.js files.

Note: This tutorial may require a newer version of BotSDK files. If the downloaded BotKit SDK files do not contain LiveChat.js and LiveChatAPI.js files, contact Kore.ai Support.

Edit config.json

In your BotKit SDK folder, you will need to edit the config.json file that contains the connection information to your callback server. For this tutorial, you are using ngrok, along with the security credentials generated when you created the app registration for the Airlines Travel Planning assistant, and adding a new line to config.json file that contains the license key for your LiveChat agent application.

Modify the default config.json file for the following parameters as follows:

  • url – Add this parameter with the Forwarding URL from ngrok
  • apikey – Update parameter using Client Secret
  • appId – Update parameter using Client ID
  • liveagentlicense – Add this parameter with license key from LiveChat Chat URL as shown in the following example.
{
"server": {
"port": 8003
},
"app": {
"apiPrefix" : "",
"url": "< ngrok Forwarding URL >"
},
"credentials": {
"apikey": "< Client Secret from XO Platform >",
"appId": "< Client ID from XO Platform >"
},
"redis": {
"options": {
"host": "localhost",
"port": 6379
},
"available": false
},
"examples" : {
"mockServicesHost" : "http://localhost:8004"
},
"liveagentlicense": "< LiveChat License Key >"
}

The following code is an example of a fully defined sample config.json file:

{
"server": {
"port": 8003
},
"app": {
"apiPrefix" : "",
"url": "https://1b47f1d2.ngrok.io"
},
"credentials": {
"apikey": "dTTmuTOgs8SOCwXXXXXXXXE0EPy28fXojOJWedOEF6s=",
"appId": "cs-6e5eb411-XXXX-XXXX-9ba2-ccc748d4f719"
},
"redis": {
"options": {
"host": "localhost",
"port": 6379
},
"available": false
},
"examples" : {
"mockServicesHost" : "http://localhost:8004"
},
"liveagentlicense": "89XXX15"
}

LiveChat.js and LiveChatAPI.js

The LiveChat.js and LiveChatAPI.js files in the BotKit SDK contain the JavaScript that controls the seamless transfer from the dialog task in the assistant to an agent application. For this tutorial, you will configure the dialog transfer using LiveChat with the following information:

In XO Platform, select the Airlines Travel Planning assistant, and then on the Configurations tab, click the Expand icon for the General Settings section. The General Settings section is displayed. Copy the:

  • Bot Name – The name of the assistant on the XO Platform, for example, Airlines Travel Planning_KORE ADMIN_10.
  • Bot ID – The XO Platform system has a unique ID for the assistant, for example, st-26cfae3a-XXXX-XXXX-991a-376b7fe579d5.

In both the LiveChat.js and LiveChatAPI.js, update and save the following case-sensitive variables with the settings from your assistant’s instance:

var botId = “< Bot ID >”;

var botName = “< Bot Name >“;

For example:

var botId = "st-26cfae3a-XXXX-XXXX-991a-376b7fe579d5";
var botName = "Airlines Travel Planning_KORE ADMIN_10";

Starting the BotKit SDK Application

To start using the Airlines Travel Planning assistant with Agent Transfer using the BotKit SDK, node.js must be installed and ngrok running listening on port 8003.

To start BotKit SDK

In a Terminal window, navigate to your BotKit SDK folder, and then enter the following command:

node app.js

When node.js is running, you should get the following response:

app listening at http://:::8003

Testing the BotKit SDK Bot

To test your assistant, you need the following:

  • ngrok running – Command = ngrok http 8003
  • node.js running – Command = node app.js
  • LiveChat – Logged in and web Chat page open
  • XO Platform – Start Talk to the Bot by clicking the following icon on the bottom right corner of any XO Platform page:

After the Agent Transfer dialog is started, messages are exchanged between LiveChat and the BotKit SDK.

Next Steps

To use the Kore.ai assistant in your application, you would need to configure the BotKit SDK and register your application with CallBack URL to point to your company application server instead of ngrok. And of course, to share the assistant with other users, the Enterprise Admin for your company would need to approve the assistant, and then deploy it to those users after you publish.

メニュー