Let Oracle AI help you with your next likely actions
There might be many articles and presentations available about various releases and updates related to cloud readiness from Oracle and now are shortly approaching 23B. However, I have some interesting observations that I would like to share which were part of 23A updates under Financials. The new feature is about getting a news feed which will help the user(s) for their probable next action which needs to be taken, this new feature is known as ‘Get suggestions on your next likely actions’.
The activities that each user is most likely to do during a session are predicted and suggested using artificial intelligence and machine learning algorithms. The algorithms analyse historical navigation habits and behaviour of each user’s and users like them to predict which navigation target they are most likely to visit.
Oracle Fusion Cloud Apps offer a broad variety of capabilities, making it occasionally challenging to find navigation to a certain activity. For instance, if reviewing and approving expenditure reports is a duty that managers in your company frequently complete, new managers will get a suggestion in their news feed to complete that activity. Other team members may see this assignment as a suggestion in their news feed if a few team members are entering performance objectives.
As can be seen in the above screenshot the ‘Suggestions’ tab lays out six important recommendations for the user for which an action may be necessary. The “promote” and “dismiss” buttons are also displayed in the box along with the explanation for the action. The dismiss button will stop the action from ever showing up in that user’s Suggestion tab in the future, while the promote button gives the machine learning algorithms positive feedback.
The Recommendations can alternatively be ranked in the order in which they are shown. System administrators have the ability to rank, elevate, and/or deactivate activities to control the order in which they show in the Suggestions stream.
In the above example, the left field demonstrates how the assigned rank is established, and the right field demonstrates how an administrator made a display on a specific user’s Recommendations page. The top of the list on the right is Opportunities Marketplace, which is ranked 1 in this instance. The position available will be filled by proposed actions generated by ML.
The administrator can also use a group of people (users) as a segment and assign a Suggestion across segments. For instance, a segment ‘Sales Representatives’ needs to take an action on the 1st Monday of the month. In this case, a Suggestion set can be assigned to the segment ‘Sales Representatives’ which will be displayed on the first Monday of each month.
The below screenshots provide how to create a segment, create a suggestion set and rank the same so that the segment users have the activity which is ranked 1st in their Suggestion news field.
In the above example, the segment ‘Sales Representatives’ needs to raise their respective expense reports on 1st Monday of each month and that is ranked first and visible on the top right, whereas other actions will be predicted by ML algorithms.
There are also reports available to measure the usage of this functionality such as the Engagement Report and the Accuracy Report. The Engagement Report shows the percentage of selected users who clicked suggestions displayed to them over a particular period. For instance, the number of new hires during a specific time frame clicked on displayed suggestion.
The Accuracy Report shows the percentage of users in the region/country or location who visited suggestion targets, for example if a group or segment of people in HQ frequents the suggestion(s)- that means that the suggestions are accurate. Higher the percentage of people, better is our algorithm.
As an organization’s activity is quite dynamic, there are many new people joining across various business divisions/segment, this functionality will help a quick turnaround of activities with less amount of time dedicated in training which can be utilised in the core activity which the role demands. I believe that organizations should seriously consider this option, on pilot basis to start with then may be at a large scale.
News feed Suggestions feature of Oracle Cloud Applications across ERP, HCM and CX. To enable this feature, you need to log a Service Request (SR) through My Oracle Support and request for a promotion code. This feature is currently in Controlled Availability and requires a promotion code. For details, see PRC:PO: How to Apply for News Feed Suggestions Under Controlled Availability (Doc ID 2915145.1). This functionality is part of the fee structure, no additional fees/payments required.
About the Author:
Gaurav Oza is an Oracle Cloud/eBS P2P Functional Support Consultant here at Version 1.
