Paris - Conversational API Framework

#Paris Release , #Virtual Agent , #Conversational API Framework

Brad Tilton

3 minute read

When I was looking through the content of our Paris release the Conversational custom chat integrations framework, while being a mouthful, jumped out at me. I’ve talked to numerous developers over the past year or two about our Virtual Agent and one of the most common questions is something along the lines of: Can I interact with ServiceNow’s Virtual Agent through a different chat tool outside of the OOB Slack, Teams, and Facebook integrations? I can finally answer yes!

Conversational Custom Chat Integrations Framework

This new capability in Paris is a framework of scriptable APIs that enable customers to plug into ServiceNow’s Virtual Agent through any chat interface of their choice. This includes any channel that supports a conversational interface like a web portal, custom mobile app, Whatsapp, SMS, etc. (voice channels are not currently supported). It’s important to note that this is not an API you can interact with in itself, but a framework you can use to build an integration.

When you enable the Conversational Custom Chat Integration plugin (com.glide.cs.custom.adapter), you can use the the framework to transform a message from an external chat client to the Virtual Agent Chat Server (VACS), and then back from VACS to the external client as shown below.

2020-09-08-22-53-26.png

You start by creating an integration through a scripted REST API and then use Flow Designer action scripts to do the inbound and outbound transformations of the message content so that both the VACS and external chat client can communicate.

Use Cases

There are two primary use cases at the moment for this new framework:

  1. Bringing chatbot capability to enterprise and 3rd-party chat interfaces. This enables customers to leverage VA in new and emerging chat interfaces and channels using a standardized method of building and utilizing the Virtual Agent.

  2. Bot-to-bot integration that allows Virtual Agent to be used as a child chatbot to a parent chatbot like IBM Watson, MS LUIS, or Google Dialog flow that does orchestration across chatbots.

Try it out!

Looking for something a little more in depth than this post? If you want to learn a little more about the Conversational Custom Chat Integrations framework I highly encourage you to take a look at the Telegram Messenger Share App. The app has some starter files and then the accompanying documentation is basically a step by step lab guide on building out a simple Telegram bot and integrating it with Virtual Agent using the new custom adapter framework.

To see a complete list of virtual and live agent chat features supported by this new framework, visit the docs site.

Let us know what kind of conversational integration you’re planning to build in the comments below!

Comments