Agent Chat Available Only for Specific Applications or Services

[nmaneely]

If a Service Desk only wanted to provide Agent Chat for specific Applications or Services, what would be the most efficient way to do this as I don't believe it is possible to build this into the AWA Routing conditions for the Chat Queue?

[Eliza]

Hi @nmaneely,

When you say applications or services, do you mean routing work to agents only if they relate to a specific topic?

If this is the case, then you can look at using Context based routing, as documented in the video here. You can skip to the 12:50 mark to see a demo.

In this scenario, you can ask users what their request pertains to prior to routing them, so you can either decline to route them to a live agent if you don't support that service (whilst providing alternate means for resolution such as linking to the knowledge base) , or use the user's answer to route them to the correct agent.

Let me know if you have further questions!

[nmaneely]

The Service Desk is only able to provide Chat for End User Applications/Services so we want to be able to, based on the application/service the user is having trouble with, either offer them the ability to chat with an Agent or just create an Incident. To complicate it further, we are using a 3rd party virtual assistant (not Virtual Agent) so just wanted to confirm if context based routing (which is what we are currently implementing) is the best/only way to solve for this.

[Eliza]

Yes - as far as I know using context variables is the best way to implement this functionality. I imagine the flow is as follows:

1. User opens up 3rd party virtual assistant

2. Virtual assistant asks them which application/service they need help with

3a. If they select one of the options for which live agent support is available, then a context variable is set with that application/service name.  That context variable is then used to route them to a live agent for that specific application/service.

3b. If they select one of the options which require an incident to be logged, then they will be prompted to create an incident either via the virtual assistant, or directed to an incident creation form.

Are you running into any issues/pain points with implementing this?

[nmaneely]

Eliza,

The flow is pretty much as you describe. One issue we have is the ability to hide or obfuscate the Agent Chat URL. As Chat is not offered for all Apps/Services, we do not want users to save the URL and then be able to directly connect back into the Chat Queue to start a new conversation without having been routed through the virtual assistant to determine if chat is provided for the issue they are encountering.

Any thoughts/suggestions?