Whats New in Oracle Intelligent Advisor 20A?
The crop of new features in 20A this month can also be filled out with some extra new features that crept into 19D when it was updated about a month after the initial release.
In 20A, the focus is very much on enhancing the connection with Oracle Engagement Cloud. As those who work with it know, up until now getting set up with an Intelligent Advisor interview has required Groovy Script. More importantly, Oracle Engagement Cloud / Oracle Intelligent Advisor integration has lagged behind the others (notably the Oracle Service Cloud, or even Oracle Siebel integrations) in terms of functionality. And finally, in the past it was absolutely awful in performance terms, notably the GetMetaData was renowned, at least where I was working, for taking up to 4 minutes to provide a response.
The performance issues were worked on a while ago, and Oracle have made great strides in that direction, so now it is fantastic to see that the other aspects of the connector are getting some love too:
Dynamic reference data loading – the ability to load in additional data from Engagement Cloud (for example, product catalog or transaction history information) after the user is already part way through an advice experience, Yup, ExecuteQuery comes to Engagement Cloud integrations.
Native Intelligent Advisor control in Application Composer – adding an Intelligent Advisor interview into a subtab of an agent workspace no longer requires groovy script. Essentially a “plug and play” drop-in component for your Oracle Engagement Cloud subtab. You can read more about it here.
Connector support for the Case object – Given the broad reach of Oracle Intelligent Advisor (benefits, law enforcement and so on) in the Public Sector and other “case focused” industry use cases, Case records can now be loaded, updated and created directly from Intelligent Advisor interviews. This includes support for child objects of case (contact, household, message, resource and custom child) within the same interview. For further details, you can click here.
getImageSource()– Returns the URL of the image to be displayed
getLinkUrl()– (Optional) Returns the URL of the link the user should be navigated to when they click on the image
openLinkInNewWindow()– Returns true if the image’s link URL should be opened in a new window
getHorizontalAlignment()– Specifies the horizontal alignment of the control
getWidth()– Returns the width of the control in pixels
getHeight()– Returns the height of the control in pixels
getCaption()– Returns the image description (alternate text)
If you want to learn about Control Extensions, you can read the book.
There are a few other enhancements, notably a versioned authentication in REST. But the OIA team have once again moved the bar higher and I think they deserve a big round of applause.
4 thoughts on “Whats New in Oracle Intelligent Advisor 20A?”
We have a requirement where we need to integrate Oracle Digital Assistant with Intelligent Advisor/OPA.
Can we integrate by implementing custom web-service connector framework?
Hey Deepthi. Actually you shouldn’t need to do that. You can use the Chat API (enable it on your deployed project) as the basis for your integration. Basically you can set up a new skill in digital assistant and use Intelligent Advisor to provide the content of the skill. The Chat API provides a REST-based question-by-question interface with some simple navigation capability. There is an excellent overview in the online help https://docs.oracle.com/en/solutions/policy-automation-with-chat-application/index.html#GUID-1EC5C1AB-2545-4E44-A4AE-6F028FA45361 and even an example custom component to connect OPA to ODA built by my friend and amazingly skilled Phil W https://github.com/oracle/opa-oda-chat
Thank you for your valuable suggestions. This integration they have already in place ( Integrating OPA with ODA-chatbots), but they want vice-versa like they want to integrate ODA with OPA. Is it achievable by implementing custom web-service connector framework or do we need to use OPA REST API calls?
Hi. Could you give me a use case? As always it’s hard to visualise without details