Supported Virtual Agent and Agent Chat features for custom chat integrations

The chat integrations framework supports a multitude of Virtual Agent and Agent Chat features to provide a robust conversational experience.

User authentication, authorization, and secure connection setup

Synchronous authentication of messages

Provider authentication is synchronous. They will get an immediate response with the request successful or failure code. For configuration information, see Synchronous authentication of messages.

Note:

As of the Paris release, asynchronous authentication is no longer supported.

Account auto-linking

Allow a user's third-party account to be auto-linked to their ServiceNow® profile. This feature bypasses the prompt for the user to link their accounts manually. Without account linking, conversations will take place in Guest mode. For configuration information, see Account linking in pre-built messaging integrations.

Specify the outbound endpoint URL to which the Virtual Agent responses are posted and configure outbound authentication. For configuration information, see Outbound authentication using a token.

Secure data transfer

Sensitive data such as passwords can be masked and transferred via the integration framework securely.

Trusted media domains

Use trusted media domains to check the URLs of attachment files before uploading them. For configuration information, see Set up trusted media domains for secure file upload.

Rich control transformation

Design a rich and unique conversational experience for a channel through rich control transformation via inbound and outbound transformation scripts. All standard rich controls are supported. For detailed information about Virtual Agent controls, see the following sections:

• Assistant Designer user input controls
• Assistant Designer bot responses
• Assistant Designer utilities

Supported Virtual Agent chat features

No configuration or scripting required to implement these features in Custom Chat Integration Framework.

Support for all standard Virtual Agent controls

• Pass context variables, system parameters, or a user's time-zone along with user input.
• By default the Virtual Agent processes only one user request at a time. If a user sends a second request before the first request is finished, the second request is ignored. This does not apply when the user is connected to a live agent.
• Ignore duplicate requests from client.

Topic discovery using keywords or Natural Language Understanding

Custom chat integrations can avail topic discovery via keywords or NLU depending on how Virtual Agent is set up on your instance.

Standard chat and long-running conversations

Channels can be set up as either chat or messaging. For more information on long-running conversations, see Using Asynchronous Chat in messaging channels.

Custom Virtual Agent chat experiences

Define a custom chat experience for a specific context in which your users run Virtual Agent.

Dynamic translation

Display a message in chat client that messages are being translated if dynamic translation is enabled.

Customize Virtual Agent and Agent Chat system messages

System messages are a set of pre-built messages that can be served as bot response in common conversation scenarios such as greeting a user or transferring to a live agent. Admins can override the message text to suit their business needs.

Notifications

Deliver both simple and actionable notifications via the integration framework.

Standard timeouts for idle agent conversations

You can change the default timeout period for closing Virtual Agent and live agent conversations by adding the system property com.glide.cs.conversation_idle_timeout.

Virtual Agent interaction records

The channel (sys_cs_channel.Name), channel type (sys_cs_channel.Type), provider (sys_cs_provider.Name), provider identifier name (sys_cs_provider_application.Name) and provider channel id (sys_cs_provider_application.InboundID) fields are available for reporting.

API rate limiting

Ability to control the rate at which API calls are made via the platform rate limiting feature.

API versioning

Specify the version of the provider. The framework is able to send the expected inputs to actions and subflows and handle the expected outputs. See Configure a provider for your custom chat integration for more information.

Specify the default portal for URL navigation at the Provider Channel Identities level

For a provider channel, specify the portal in which chat links, including links in output cards and cards for Virtual Agent notifications and AI Search results, are opened. For details on setting the default portal in the Provider Channel Identities [sys_cs_provider_application] table, see Create a channel identifier for your custom chat integration. To learn more about URL navigation, see Configure URL navigation for chat links.

Supported Agent Chat features

Support for all standard live agent controls

• Transfer to live agent from Virtual Agent or directly by user
• Persistent messages after live agent is initiated
• Interaction history (chat transcripts) for live agents
• Channel-level flag to initiate conversations with live agents, bypassing Virtual Agent

Live agent status for end users

• Agent availability
• Wait time for next agent
• Display agent name and avatar, if enabled in Agent Chat settings.

Channel-level flag to initiate conversations with live agents, bypassing Virtual Agent

Enable the Live Agent Only option in the Messaging Channels [sys_cs_channel] table to route conversations directly to a live agent and skip the Virtual Agent conversation greeting. If an agent isn't available, the message No agent available is displayed to the user. For channel configuration information, see Create a new channel for your custom chat integration.

Pass the agent avatar and name if enabled in Agent Chat settings

When the Agent Names and Avatars field is enabled in Agent Chat settings, the Custom Chat Integration Framework supports the display of the agent name and avatar in the chat. For more information, see Setting up Agent Chat.

Pre-chat and post-chat surveys

After an agent ends a chat, the post-chat survey runs automatically when the following criteria are met:

• The Virtual Agent Feedback setup topic is enabled.
• The chat service channel and the queues that your live agents support are configured in Advanced Work Assignment.

• Synchronous authentication of messages
  Inbound message authentication is performed before the message is added to the hybrid queue.
• Outbound authentication using a token
  Pass an outbound token from the authentication API to the sender script.
• Message delivery order
  Control message delivery order by enabling the Maintain message ordering option in the Provider Channel [sys_cs_provider] table. You can view the message order and status in the Custom Adapter Message [sys_cs_ca_message] table.
• Virtual Agent action scripts
  You need to write a set of mandatory Workflow Studio action and subflow scripts to pass the user data to VACS, apply transformations and send the response back to your chat interface endpoint.
• Set up trusted media domains for secure file upload
  Virtual Agent uses trusted media domains to check the URLs of attachment files before uploading them. If a URL is not from a trusted domain, Virtual Agent will not upload the attachment.
• Create bot messages
  Create channel-specific bot messages.
• Integrate custom controls
  Implement and leverage custom controls built using the custom control framework in your integration.
• Multi-step responses
  An example of multi-step response is a set of images in a carousel or a date control.
• Multi-response output controls
  Implement controls that involve multiple steps to get user input.
• Send previous control to client
  Provide the ability to send previous control to client.
• Implement a skip option
  Use a rich control to provide the option for users to skip a question.
• Pass branding to a chat client
  Pass custom branding to your custom chat integration.
• Contextual actions for custom chat integrations
  Commands with a pre-defined behavior can be mapped to certain user-friendly keywords and passed as user input. This input is mapped to supported contextual actions for the most appropriate response in the chat.
• Typing indicators
  Display a typing indicator when a bot or live agent is preparing a response to a requester.