Posted: 17 Jan 2018 17:44 EST Last activity: 7 Feb 2018 14:07 EST
Intelligent Virtual Assistant support for Aspect Chat /Avaya Chat
IVA documentation says there is in built support for Alexa, FB, SMS and Email. As per this article @ https://pdn.pega.com/intelligent-virtual-assistant-archive ,
"Pega also offers an extendable architecture so that customers can build a custom integration to other conversational channels."
Has anyone extended this capability to Chat solutions provided by other vendors like Aspect & Avaya?
Is there any document reference for that ?
@GOWTHAMALLU It is absolutely possible for customers to create their own channel. The process, however, could be a bit daunting as it should be because you're creating an entirely new integration between provider and platform.
At the outset, the first thing you need to do is create classes where your configuration and interaction cases live. (all rules preferably live in <ChannelName>Channel ruleset like FacebookChannel)
Data-Channel-Configuration-<channel name> is used to store configuration and extensions for your UI like access tokens that the provider exposes.
Work-Channel-Interaction-<channel name> is used to store interaction of the user until the timeout expires.
Then, you'll need to create endpoint Services and Connectors for back and forth between your bot and the platform.
You'll also need to set up authentication so that incoming user can be mapped to an existing operator or create new operator and call pxChannelAuth activity.
pxProcess is the meat of MCP. Before calling you need to prepare input for it and after it ends you can use output to build text that your bot understands.
You'll need to map incoming message into pyInteractionCase.pyInboundMessage
The output of pxProcess is pyInteractionCase.pyOutboundMessage() pagelist which you can use to construct string (like JSON in FB) that channel provider understands.
You can get FB component from Exchange and see how you'd go onto modify behavior for your channel. Please post here if you have issues.