Being in the software industry for over 7 years I have noticed that everyone knows what an ERP (Enterprise Resource Planning) software system is. If those users don’t know the term ERP I will ask what is your financial system and then they will know what they are using. The other trend is every company in 2021 is using some type of software for a financial system but not every company has the same value on ECM software. What is ECM software and why is it important?
By Jack Shales — Sales Engineer at REQUORDIT
JD Edwards is a powerful ERP being used in a wide range of industries (Healthcare, Banking, Manufacturing, Retail, Higher Ed, Construction and more) with the goal of boosting user productivity in a flexible and scalable solution. At REQUORDIT, we strive to maximize human potential through process automation and, while we believe that JDE is a best-in-class ERP system that does indeed provide a boost to Accounts Payable organizations, we also believe that we have the tools to take your process to the next level.
When you think about your current procure to pay process with JD Edwards it is very likely that it begins with the requisition of a purchase order. The PO is most likely being manually entered by a purchasing agent in your JDE client of choice and then distributed using the limited toolset available, possibly even including several emails back and forth with vendors or needing to navigate to several different JDE screens if, for example, a PO is being cut to a new vendor that needs to be added to the address book. This process can take a lot of time away from the purchasing agents that could be better spent on value added tasks.
Although there are some out of the box tools to leverage in JDE with regards to automated notifications and other workflow capabilities I have seen firsthand the benefit of striving for further automation with a robust workflow engine like OnBase. OnBase would fill in every gap I highlighted above with its ability to send automated notifications for the currently manual vendor follow up and internal alerts as well as the ability to create responsive electronic forms that would allow the multiple data entry screens in JDE to be combined in one view.
Imagine pulling up a purchase order requisition form on a tablet or mobile device with dropdown fields integrated with your JDE data, making it easy to fill out and hard to make mistakes. If the PO is with a new vendor, the form can dynamically adjust to display vendor creation fields. After the form is submitted it can go through an approval process with dynamically assigned approvers based on the entered data or automatically create the Vendor and PO in JDE.
After the PO is requisitioned and approved the invoices soon start flowing in from the vendor. These can come in a variety of formats both paper and digital, potentially with other supporting documents that may need indexing and evaluation as delivery tickets. Again, this can create a sizable bottleneck as invoices may need to be manually opened from email, manually downloaded, then manually uploaded into JDE with all invoice data hand keyed.
In one of my recent blogs I talked about how OCR (Optical Character Recognition) is here to stay and the situation above represents the perfect use case for introducing automation with OCR, Classification, Data Capture and Machine Learning.
It was not that long ago that when we talked about OCR and Data Capture we were talking about building a pre-defined template for every single style of invoice. This not only meant one template per vendor, but potentially multiple per vendor. Building and managing these templates used to be a full-time job, but now with the advent of Machine Learning it is easier than ever to automate your inbound indexing.
Imagine a world where a vendor sends invoices as attachments to your company email then the invoices are automatically read by OCR, identified by Classification, indexed by Data Capture, then presents the user with only fields that it could not validate in JDE or could not read due to image quality. Now imagine this system getting better and better over time as you train it by correcting its mistakes. This is no longer a future state and we have many customers utilizing technologies like CloudOCR to accomplish this today.
With capture rates that can easily clear 90% after a month of invoicing this is one of the best ways to maximize the potential of your organization today and get out of the business of data entry.
I made brief mention of approvals in the purchase order discussion above. These approvals are not limited to purchase orders. In fact, any document you can think of can have its own approval process that can be dynamically structured based on one or many criteria.
For example, a $10,000 invoice for parking lot materials comes into the XYZ Construction Company AP department through email where it is automatically picked up by CloudOCR, indexed, and returned. With our workflow, this invoice will immediately go through several evaluations — Does this PO exist? Is the correct Vendor applied? Is there a potential duplicate invoice? — and more in the blink of an eye. If any data crucial to determining the approval path is not there it can route to a purpose built exception queue.
If the invoice comes in and it passes all of those evaluations it can go into a load-balanced approval queue and the next approver can be automatically notified and approve from Outlook or a mobile application. We can even performing an automatic 3-way match if all required data is captured and validated bypassing approval altogether.
Since this is an invoice for $10,000, after hitting the approval queue the workflow would check for any approvals that apply based on the indexed fields. In this case that total amount is driving what approvals need to occur, so at $10,000 it might go through two levels, needing review by the Project Engineer and Project Manager. If, however, an invoice for a $1,000,000 building comes in it might need to go through four levels of approval (Project Engineer, PM, VP of Finance, CFO). The best part is that creating new approval structures will not take the involvement of IT and can be managed by the business process owners. These approvals can also be driven by other data points such as Job numbers or combinations of values.
