Get AI Suggested Invoice Classification Rules

Hello, and welcome to today's demonstration on Oracle Fusion Cloud Sustainability. In this session, we will explore the new feature introduced in release 25B, AI-suggested invoice classification rules. This feature is designed to enhance efficiency and improve the accuracy of invoice classification within your sustainability processes.

To provide some context, Oracle Sustainability uses invoice classification rules to automatically generate activities based on eligible distributions of payables invoices. Prior to this update, when an invoice distribution did not match any existing classification rules, Oracle Sustainability would create an activity based on a default activity type defined in a profile.

These activities would then require manual review and classification before they could be posted to the sustainability ledger. At this point, I'd like to introduce a key concept, classification coverage. This refers to the percentage of invoice distributions that are classified correctly without requiring manual intervention.

Once you have set up an initial set of invoice classification rules, improving this classification coverage becomes increasingly challenging. Typically, this would involve manually analyzing activity and invoice data to identify patterns that could be used to create additional classification rules. This process, while effective, can be time-consuming and resource-intensive.

To address this challenge, Oracle Sustainability now integrates artificial intelligence to automatically identify hidden patterns in your historical activity data and payable invoice data. Based on these insights, the system can suggest new classification rules. By accepting these AI-suggested rules, you can significantly improve classification coverage, ensuring that future invoices are automatically classified without requiring manual review.

Now, let's take a closer look at how this functionality works. I'm currently on the Invoice Classification Rules page. If you choose to opt into this feature, you can generate rule recommendations by selecting Suggest Rules from the Actions menu on this page.

Clicking on Suggest Rules will initiate the Suggest Classification Rules Schedule process. This process will begin analyzing all activities that have been posted to your ledger. Depending on the volume of data, this process may take some time to complete.

Let me take the action now. Yep, the job has been scheduled. If sufficient activities from payables invoices have already been posted to the ledger, the system will analyze those and identify patterns that could be used to suggest new classification rules.

If there are not enough activities, the system will review your historical payable invoice data. It will then use generative AI to predict the likely classification category for each activity and leverage those predictions to identify patterns for rule suggestions. Once the process completes, the suggested rules will be displayed here on the Invoice Classification Rules page.

We are now back on the Invoice Classification Rules page after the scheduled process has been completed. As you can see, several AI-suggested rules have been generated. By default, Oracle Sustainability places these suggested rules in draft status. This means they will not have any impact on the classification process until you take further action. You have three options for how to proceed with the suggested rules.

Edit: if you find the rule useful but wish to make adjustments, you can select Edit to modify the rule before accepting it. Accept: if you are satisfied with the suggested rule, you can select Accept to activate it. The status of the rule will then change from Draft to Active. Reject: if the suggested rule role does not seem relevant or useful, you can select Reject to remove the rule from the system. If you reject a rule, it will not be recommended again in the future.

Once a rule is active, it functions just like any other classification rule in the system. The next time the generate sustainability activities from payables invoices process runs, it will automatically classify any invoices that match the active rule, thus reducing the need for manual intervention.

That concludes the demonstration. Today, we have explored how AI can help identify hidden patterns in your activity and payables invoice data and how these patterns can be leveraged to automatically suggest classification rules. By accepting and activating these suggestions, you can significantly improve your classification coverage and reduce manual effort in the invoice classification process. Thank you for your attention. We hope you found this session informative and valuable as you continue to streamline your sustainability processes with Oracle Fusion Cloud.