Unified Admin Guide

Select a time zone that best suits your servers' location.  

Choose the languages that the system should support, for instance, for rendering the UI.

For now, the only supported languages include:

• English

• French

• Spanish

• Italian 

• German 

This setting is currently only used in Web Widget → Language settings to show a list of supported languages for the Customer Widget UI. 

For the other components such as AgentDesk, the UI translation will be available in future updates.

Based on the supported languages chosen from the dropdown above, choose here the language which should act as a default language. For now, this is set to 'English' by default.

Specify a name to be given to the bot. 

Enabled only if the Bot Type is Rasa or Custom. Provide the URL where the default bot connector is running.

Example: http://server-ip:30800

Enabled only if the Bot Type is DialogFlow.

Give a name to the MRD.

Provide an optional description of the MRD.

Specify the maximum number of task requests that any agent can handle on this MRD. When this count reaches, the agent's MRD state will turn to BUSY and the agent will no longer receive any new tasks unless an active task is closed.

This works in combination with the Agent-MRD tasks limit. See more in Step 8. 

Note that this limit should be greater than or equal to the task limit defined for one agent on this MRD. Therefore, if the maximum tasks count is later modified to make it lesser than an agent's individual tasks limit (during Agent-MRD mapping), an error would be thrown. 

Reserved for future.

When available, this will allow us to turn the flag to ON if the requests on this MRD are considered to be interruptible. This means that the system may interrupt the agent on this MRD (during an active conversation) to route him to a conversation from a different MRD if required.

Give a name to the Routing Attribute.

Provide an optional description here.

Choose from one of the following types:

• Proficiency: Specify a default proficiency on a scale from 1 to 10.

• Boolean: Select the default value, either true or false.

Specify a default value based on the type of attribute. This default value is automatically assigned to an agent when a Routing Attribute is assigned to the agent. You can overwrite the default value at the time of attribute assignment.

Specify a queue name. 

Choose an MRD you created in step 2. 

This is when the agent has to respond to a customer message.

Type one of the following numbers in the field: 

• 1

• 2

• 3

See more details about these in Key Reporting Concepts

This is the service level threshold in seconds. This determines the time until when all requests being enqueued should ideally be picked up from the queue and routed to agents to be answered. 

This is set by the business as per their business policies and SLA reporting requirements. If a request gets answered within this time, it is considered to be answered within SL. 

See more on  Key Reporting Concepts

This is the minimum estimated wait time in seconds that will be displayed to the customer if the estimated wait time is below this number. The minimum value should be lesser than the maximum value.
Minimum Number Limit: 0 or NULL
Maximum Number Limit: 2147483647

This is the maximum estimated wait time in seconds that will be displayed to the customer if the estimated wait time is more than this number. The maximum value should be greater than the minimum value.
Minimum Number Limit: 0 or NULL
Maximum Number Limit: 2147483647

Mandatory

Optional

This is the timeout for this step. It determines the time the system waits before jumping to the next step when no agent is found. 

Add a logical expression for the Routing Engine to evaluate and identify an agent matching the criteria.

Add the small + icon to add another term in the expression. Click Add Expression to add another expression to the same step. 

For instance, the expressions added in the illustration below states: 

Any agent having (English== 8 AND Sales == true) OR (Science >= 5 AND Sales == true), is best suited to answer the request being enqueued to this queue.

Provide a name for the channel provider such as Dialog 360 for WhatsApp.

This setting is used to determine which channel types this channel provider will  support. For instance, you could configure one channel provider to support both WhatsApp and Facebook messages. So one channel provider can be used to serve both types of channels.

Based on the supported channel types, you can add corresponding channel connectors and use them while configuring new channels.

Give a webhook URL for the Channel Provider where this channel provider will listen to the events:

• Recommended to use the service name of the component in the Channel Provider > Provider Webhook field. However, FQDN can also be used with some additional custom configurations.

• Provide the service name of the Connector in the Provider Webhook field as http://{SERVICE-NAME}:{SERVICE_PORT}. Replace the {SERVICE_NAME} and {SERVICE-PORT} in the URL with the component’s k3s-based service name and port (it can be fetched using the following command: kubectl get svc -n expertflow )

The provider webhook URL is given in the following format:

For Web provider

Kubernetes based: http://ef-web-channel-manager-svc:7000

HELM based: http://ef-cx-web-channel-manager-svc:7000

For WhatsApp

Kubernetes based: http://ef-whatsapp-connector-svc:8080

HELM based: http://cx-channels-whatsapp-connector-svc:8080

For Facebook

Kubernetes based: http://ef-facebook-connector-svc:8080

HELM based: http://cx-channels-facebook-connector-svc:8080

This is the name of the channel connector.

Select the desired Channel Provider (created in the above step) to configure this channel connector for the channel provider. Based on the selected channel provider, you'll see the corresponding configuration attributes to be specified for this channel connector. See the illustration below.

This shows the list of other custom attributes created while defining the Channel Provider you selected above. Available only if some custom attributes were added while defining the Channel Provider in the above step. 

For example, for the default Dialog 360, 

HOST-URL

API-KEY

For details on Dialog 360 connector configurations, see 360-Connector Configuration Guide for WhatsApp

Specify a channel name. 

This is the service identifier of the service that the customer dialed or reached. It could be a DN (Dialed Number) such as a WhatsApp number, any website URL, Facebook ID, Viber ID.  

This is the Bot Connector that was added under Bot Settings. 

You need to choose a bot connector here regardless of the Channel Mode. This is because a bot is an integral part of a conversation who acts as an authoring party, monitors the conversation events, identifies intents, and takes appropriate actions (for instance, closes a chat if the customer activity timeout reaches). 

This is the Channel Connector that should be used to send/receive messages to/from the channel. Choose the desired channel connector created in the above step to be associated with this channel.

This list is rendered based on the Channel Type.

This is the time in seconds that the system waits for the customer's response against a message sent by the agent or bot. When this time reaches, the system considers the chat as inactive. The bot closes the channel session of the customer and the agent receives Channel_Session_Ended notification within the conversation. The message composer on the agent side is greyed out to disable sending messages.

Choose a channel mode. Possible values are:

• Hybrid: This the default mode of communication where a bot is always there to assist agents in the background (even if not conversating with the customer directly). This is the only available mode as of today. 

• Bot: Choose this option for self-service, Bot-only conversations. Reserved for future use.

• Agent: Choose this for requests to be routed to agents. Reserved for future use.

This determines if the channel needs to push incoming requests to a queue to be routed to agents, or, broadcast to a Pull-based List to be taken up by any agents subscribed to the List.