Jump to: navigation, search

Deploying Chat Bots

Genesys Intelligent Automation can provide your customers with chat bot-based access to your WebIVR applications. Customers can converse with a bot using natural language to accomplish various business needs, such as making a payment or checking an account balance.

Warning
Please note that the customer is responsible for ensuring that the environment and bot applications they build are properly configured and secured according to PII and HIPAA requirements.

PureConnect

For information on deploying PureConnect chatbots, see PureConnect with Genesys Intelligent Automation for more information.

Genesys Engage

The following components are required:

Important
For SMS (Short Message Service) functionality, you must also install Social Messaging Server 8.5.x with SMS plugin.
Important
To test your bot functionality using Virtual Call, please set the Channel field to WEBCHAT when starting your Virtual Call session..

Once installed, the bot acts similarly to a WebIVR application using prompts from the associated visual persona. However, if using SMS, remember that customers cannot use special characters in response to queries from the bot.

Preparing the chatbot environment

Install Digital Messaging Server

  1. Refer to the Digital Messaging Server Guide for information on how to install Digital Messaging Server.
  2. Copy the chatbot jar file to the Digital Messaging Server subfolder for chatbots. For example: C:\Program Files\GCTI\eServices 9.0\Digital Messaging Server\DMS\media-channel-drivers\channel-chatbot\bots-repo\.

Use Genesys Administrator to configure Digital Messaging Server

  1. Open Genesys Administrator and log in to your configuration environment.
  2. Go to Provisioning > Environment > Applications and open the Digital Messaging Server object.
  3. Create a section called channel-chatbot-monitor-bots.
  4. Create a chatbot option. You must ensure the option name exactly matches the name of the jar file that you previously uploaded to Digital Messaging Server. For example: fish-cbp-9.0.0.jar.
  5. Set the value of the chatbot option to point to the load balancer for VUI Server. You can also add the parameters downstream_request_timeout_millis and session_timeout_minutes to set timeout values. In the following example, the request timeout is 120000 milliseconds (120 seconds) and the session timeout is 30 minutes:
    { "base_url" : "http://www.yourcompany.com:8081/fish-vui", "downstream_request_timeout_millis" : 120000, "session_timeout_minutes" : 30}
  6. Click Save & Close.

(Optional) Use Genesys Administrator to configure SMS Server

Important
  • Use this section only if you want to provide SMS (Short Message Service) access to your chatbot. Otherwise, go to the next section.
  • Refer to SMS Server (part of eServices) documentation for more information on SMS Server.
  • This feature uses the Session Mode of SMS Server, which refers to creating and keeping an interactive conversation between a mobile client and an agent in the form of a conventional chat session. All messages received and sent during this session are associated with one interaction, which corresponds to this SMS session.
  1. Open Genesys Administrator and log in to your configuration environment.
  2. Go to Provisioning > Environment > Applications and open the SMS Server object.
  3. Open the channel that you want to use with the bot.
  4. For the option inbound-route, specify the access point in your routing strategy that is used to place submitted interactions for incoming messages.
  5. Click Save & Close.

Use Interaction Routing Designer to configure the chatbot

Once you have installed the bot interface, you can use Interaction Routing Designer to configure settings.

  1. Open Interaction Routing Designer and log in to your environment.
  2. Open the routing strategy that is triggered by the session endpoint, and add an External service block to the start of the flow.
  3. Open the External service block and go to the General tab. Configure the settings as described below:
    • Application type - Set to the Digital Messaging Server or Social Messaging Server in your environment.
    • Application name - Set to the Digital Messaging Server or Social Messaging Server in your environment.
    • Service - Set to ChatBotPlatform.
    • Method - Set to StartBot.
    In the Parameters section, add the following:
    • Nickname - The bot name displayed to customers.
    • SiteID - The ID number of your Genesys Intelligent Automation application to use with the bot.
    • IsTestCall - If true, the test version of the application is used. Otherwise, set to false to use the production version.
    • AuthToken - The value of the Authentication Key field in your company page.
    • StopBotOnAgentArrival - Set to true if you want the bot to end once an agent joins the session. Otherwise, set to false.
    • StopBotOnCustomerLeft - Set to true if you want the bot session to end once a customer leaves the session. Otherwise, set to false.
    • Visibility - Set to ALL.
    • ChatBotID - Set to speechstorm-chatbot.
    • ChatBotName - Do not enter information in this field.
    • _umsChannel - Set to channel-chatbot.
  4. Click OK.

Genesys recommends that you add new blocks to your routing strategy, just after the External service block, to pause the strategy until the chatbot has completed its task. Otherwise, the routing strategy might queue up the session for an agent and, as soon as an agent joins, the StopBotOnAgentArrival parameter will cause the bot to terminate early.

To determine if a chatbot interaction is still alive, Genesys Intelligent Automation checks the following interaction data values:

  • IsOnline - If the value is 0, the bot is offline and you can terminate the interaction.
  • FishTransferRequested - If the value is true, you can proceed to route the interaction to an agent.

Refer to the Interaction Routing Designer documentation or consult your Genesys representative for information on the best routing strategy for your environment.

This page was last edited on September 18, 2020, at 13:40.
Comments or questions about this documentation? Contact us for support!