You are on page 1of 3

Oracle AI Apps:

Intelligent Account
Combination Defaulting
Oracle AI Apps for ERP: Intelligent Account Combination Defaulting assists the
payables function by using artificial intelligence and machine learning to default
code combination segments when processing invoices not matched to
purchase orders (POs). Predicting and defaulting code combination segments
reduces manual keystroke effort, reduces human errors, saves invoice
processing time and reduces costs.

Oracle AI Apps Benefits


ORACLE AI APPS
 Pre-built
Oracle Artificial Intelligence Apps (AI Apps) delivers powerful artificial intelligence (AI) and machine  Immediate benefits
learning (ML) functionality within Oracle Fusion Cloud applications. Oracle data scientists, software  Continuously learns
engineers, and subject domain experts develop AI Apps, utilizing advanced AI and ML, to directly
embed AI-powered outcomes within Oracle Cloud applications. The AI features are domain-
specific and pre-built, requiring no data science or machine learning expertise to use or maintain.
Inherent in machine learning is the ability to learn and adapt, therefore users enjoy immediate and
continuous benefits.

Intelligent Account Combination


INTELLIGENT ACCOUNT COMBINATION DEFAULTING Defaulting Benefits
According to measured telemetry data captured by Oracle, when processing invoices for payment,  Reduce manual key entry
one of the most key stroke intensive tasks is entering code combination segments for invoices not  Mitigate errors
matched to POs.  Save time and money

Account payable clerks often spend a considerable amount of time analyzing invoice data, looking
up relevant codes, engaging additional resources and expertise, and manually typing in code
segments. Often viewed as a monotonous chore, the effort is tedious and prone to error, resulting
in invoices on hold, long approval cycles, and expensive accounting fixes.
Oracle alleviates these issues by applying artificial intelligence and machine learning to the invoice
processing workflow to predict and default code segments, simplifying invoice processing, saving
time, reducing costs, and reducing manual key entry.
The feature is provided as standard Oracle Cloud ERP functionality, and can be enabled within
minutes without the need for system integration efforts or data science expertise.

1 DATA SHEET | Oracle AI Apps: Intelligent Account Combination Defaulting | Version [1.1]
Copyright © 2022, Oracle and/or its affiliates | Public
Intelligent Account Combination Defaulting Workflow
Intelligent Account Combination Defaulting is invoked during the invoice processing workflow.
Let’s look at an example invoice processing workflow to see how Intelligent Account Combination Defaulting works.

Figure 1. Oracle Cloud ERP invoice processing screen with defaulted Distribution Combination (code combination)

When preparing an invoice, invoice header information is either directly input through Oracle Intelligent Document
Recognition (IDR), other OCR products, electronically, or entered manually by an account payables clerk.
Similarly, the individual line(s) information for the invoice will be entered. Intelligent Account Combination Defaulting uses
the header and line information as input to the machine learning algorithms in order to predict each segment of the code
combination. For best prediction results, all relevant data in the header and line fields should be entered before the
prediction is invoked.The prediction is triggered when the payables clerk positions or tabs the cursor into the Distribution
Combination field, and the predicted segments are automatically defaulted in the field. Each individual segment is predicted
separately. A yellow box surrounds the Distribution Combination field as a visual indicator that a prediction was defaulted.
If a segment is not defaulted, or an incorrect segment is defaulted, the payment clerk will enter the missing segment, or
correct the incorrect segment. The machine learning algorithms learn from these entries, and over time will improve on
prediction throughput and accuracy.

How Intelligent Account Combination Defaulting Works

Predicted Segments

Supervised Learning Machine Learning Models

Figure 2. Data and machine learning flow for Intelligent Account Combination Defaulting

2 DATA SHEET | Oracle AI Apps: Intelligent Account Combination Defaulting | Version [1.1]
Copyright © 2022, Oracle and/or its affiliates | Public
Figure 2 illustrates at a high level how Intelligent Account Combination Defaulting uses data and machine learning to predict
the individual code combination segments when invoked during invoice processing.
At the center, depicted by the graph chart, are the machine learning models utilizing an artificial intelligence approach called
supervised learning. In supervised learning, known and correct labeled input-output data is used as examples to train the
models. Then when the models are presented with new input data, the output data can be determined from that training.
In the case of Intelligent Account Combination Defaulting, the training data is a company’s historical invoice transaction
data. The input data consists of a variety of data features, such as invoice header and line information, including supplier
name, supplier site, description, amount, and other data points. From those previous transactions, the models recognize the
patterns that determined how the final code combination segments (the data outputs) were applied. Typically, 12 to 18
months of historical invoice transaction data is ingested for model training to pick up on seasonal patterns, for example,
invoices paid quarterly or annually. However, if at least 3 months of invoice history is available, the models will perform
satisfactorily and render predictions “out-of-the-box”, without cold start issues sometimes associated with AI solutions.
Once the models are trained and enabled in a production environment, when a new invoice is being processed and the
models are invoked (by tabbing or placing the cursor in the Distribution Combination field), the models read the new invoice
information, i.e., supplier, site, amount, description, etc., and input this data into the models to create a prediction (data
output) for each code combination segment.
A goal is to accurately predict as many segments as possible, but not at the expense of making an incorrect prediction.
Therefore, the models also calculate an internal confidence score for each segment prediction. If the confidence score does
not meet a pre-determined threshold, then that particular segment prediction is not defaulted, that is, the segment is left
blank, and the user is prompted to manually enter the segment. In the unlikely event that a predicted and defaulted segment
was incorrect, the user will correct the segment.
A benefit of machine learning, is the ability for continuous learning and improved prediction accuracy. When segments are
manually input, or a defaulted segment is changed, that data is fed back to the models. In practice, the machine learning
models are continuously training, meaning the introduction of new suppliers, or even changes to a company’s accounting
structure will be picked up by the machine learning without any user intervention. Just using the system in its everyday
intended manner will render improved prediction results over time.

Implementation Considerations
Intelligent Account Combination Defaulting does not require a separate license to operate, and is included with a Oracle
Cloud ERP standard license.
Implementation requires a simple self-service setup operation. The feature can be enabled/disabled through a checkbox
within Oracle Cloud ERP Feature Selection.
Intelligent Account Combination Defaulting is currently available (as of February 2022) for early adopter customers. Contact
your Oracle sales representative for more information.

Connect with us

Call +1.800.ORACLE1 or visit oracle.com. Outside North America, find your local office at: oracle.com/contact.

blogs.oracle.com facebook.com/oracle twitter.com/oracle

Copyright © 2022, Oracle and/or its affiliates. All rights reserved. This document is Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be
provided for information purposes only, and the contents hereof are subject to change trademarks of their respective owners.
without notice. This document is not warranted to be error-free, nor subject to any other
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC
warranties or conditions, whether expressed orally or implied in law, including implied
trademarks are used under license and are trademarks or registered trademarks of SPARC
warranties and conditions of merchantability or fitness for a particular purpose. We
International, Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are trademarks or
specifically disclaim any liability with respect to this document, and no contractual
registered trademarks of Advanced Micro Devices. UNIX is a registered trademark of The Open
obligations are formed either directly or indirectly by this document. This document
Group. 0120
may not be reproduced or transmitted in any form or by any means, electronic or
mechanical, for any purpose, without our prior written permission. Disclaimer: If you are unsure whether your data sheet needs a disclaimer, read the revenue
recognition policy. If you have further questions about your content and the disclaimer
This device has not been authorized as required by the rules of the Federal
requirements, e-mail REVREC_US@oracle.com.
Communications Commission. This device is not, and may not be, offered for sale or
lease, or sold or leased, until authorization is obtained.

3 DATA SHEET | Oracle AI Apps: Intelligent Account Combination Defaulting | Version [1.1]
Copyright © 2022, Oracle and/or its affiliates | Public

You might also like