This post shares Fusion ERP Analytics sample content as part of Oracle Fusion Data Intelligence (FDI), a family of prebuilt, cloud-native analytics applications for Oracle Fusion Cloud Applications. These sample artifacts are intended to demonstrate potential use cases, accelerate implementation efforts, and provide examples of how customers can extend and customize FDI content to meet their business needs. To explore additional reusable sample content, visit the FDI Content Library.
Content Description: AI-enabled Customer Collections Risk Analysis helps collections and credit teams focus their effort on the customers and invoices most likely to pay late or default, through natural language interactions that make analytics more intuitive and accessible.
Users can ask questions conversationally and instantly see outstanding and overdue receivables by customer risk decile, how average days outstanding varies by payment terms, which invoices carry the highest default score and predicted delay, and when payment is expected. The Model Training Statistics canvas compares actual defaults with classifier predicted defaults by score decile, so users can judge how far to rely on the predictions. This workbook enables analysis by Customer, Customer Risk Decile, Default Score Decile, Transaction Class, Payment Terms, Business Unit, Ledger and Fiscal Calendar. This analysis uses Analytics Currency.
Content Type: AI Data Agent + Workbook
Application: ERP
Fusion Module: Receivables
Target Persona: Collections Manager, Collections Specialist, Credit Manager, Accounts Receivable Manager, Controller
Subject Areas: Financials – AR Aging
Business Use Case/Questions:
● What is the total outstanding amount by customer risk decile?
● Which customers in risk decile 10 have the highest overdue amount?
● What is the average invoice days outstanding by customer risk decile?
● Show overdue invoice count and overdue amount by customer with their risk decile
● Which customers have more than 50% of their outstanding amount overdue?
● What is the overdue amount by payment terms for <Business Unit>?
● How does average invoice days outstanding compare across payment terms for customers in risk deciles 9 and 10?
● Show total outstanding and overdue amounts by business unit
● What is the total outstanding amount by default score decile?
● Which invoices have the highest predicted delay days?
● Show the top 20 invoices by risk score with their predicted payment date
● Which overdue invoices are in the highest default score decile?
● What is the overall classifier accuracy percentage for the current model?
● Compare actual defaults versus classifier predicted defaults by classifier score decile
Best Practices:
● Invoices without a prediction (for example, credit memos) have no default score decile and are excluded from the Default Risk chart. Use the Prediction Processed Indicator filter to include or exclude them.
● Credit memos carry negative outstanding amounts and can offset a decile's total. Filter Transaction Class Name to Invoice for a gross risk view.
● Classifier Accuracy Percentage is predicted defaults ÷ actual defaults per score decile. Review the Model Training Statistics canvas before acting on predictions in the lowest deciles, where accuracy is lowest.
● The Ledger Name dashboard filter is passed to the AI Agent, so select the ledger first; questions comparing ledgers return only the selected ledger — this is expected behavior.
Pre-Requisites:
- Fusion Data Intelligence Release: 26.R2 or higher
- Ensure the relevant FDI application and functional areas are activated in your FDI instance
- Application: Fusion ERP Analytics
- Functional Area: Receivables
- User has permissions to import workbooks and agents
Download and Installation Steps: Available in the FDI Content Library.
We’d love to get your feedback and business use cases to help build a more sample FDI content.
