Skip to main content

«  View All Posts

How to Automate Financial Analysis in JD Edwards Using AI

August 5th, 2025

12 min read

By Nate Bushfield

 

This session highlights how the JD Edwards Digital Assistant streamlines financial operations. Led by Manuel Nera and Michael Embringham from ERP Suites, the presentation demonstrates how AI enhances financial decision-making, automates tasks, and provides valuable insights. The Digital Assistant, Franklin, helps users query financial data and generate reports using natural language, all integrated with JD Edwards. It showcases applications like income analysis, sales trends, and real-time financial insights, emphasizing the ease of use, security, and flexibility of AI-driven financial solutions.

 Ask ChatGPT

Table of Contents   


  1. Introduction
  2. AI in JD Edwards
  3. Why AI?
  4. Use Cases for Financial Operations

  5. Franklin - The Digital Assistant 
  6. Demo of Franklin's Capabilities 
  7. Conclusion and Next Steps

Transcript

Introduction

Hello, everyone. Thank you for joining today's session on streamlining financial, financial operations with the JD Edwards Digital Assistant. My name is Manuel Nera, Vice President of AI and Products at ERP Suites, and I'm joined by Michael Embringham, one of our, you know, top financial consultants with JD Edwards. So thanks for joining as well, Michael.

We're going to jump right into it because I know we have a packed day today. So we're starting Day 2 of AI week. Yesterday start off with a bang with a couple of keynotes, many sessions and we're hitting a second day here. So today's presentation is focused around talking about the digital assistant capabilities of AI and how in specifically within the financial sector, it can drive value, It can provide insights. It can be an intern as well like one of my colleagues MO at ERP suite slides to call AI to help us out with our daily tasks and routine task periodic tasks in the financial space. So with that I will advance. Michael, you still able to see my slides? Just making sure, yes, I can still see them. Great. OK, because I'm seeing a different version of it, but then just make sure everything's good. So just a little bit about ERP suites. We're an end to end consulting company for JD Edwards. Everything and anything you can think of. You know, we do applications, technical side hosting, managed services and of course we have collection of products and of course AI.


AI in JD Edwards

So let's jump into artificial intelligence as was discussed yesterday in Julie's keynote as well as the keynote that MO and I delivered, AI is quickly becoming part and parcel of what companies do.
Now, some companies may be on the beginning part of the journey, others may be a little bit more advanced, but the reality is the companies are getting value. And I'll share some more statistics today in terms that substantiate that. Secondly, AI encompasses a broad range of technologies and services. And as we saw yesterday, right, Oracle, like some of the other vendors in the enterprise AI space, have a collection of services that are focused around a variety of different capabilities, whether it's anomaly detection or whether it's bringing in and augmenting the learning of your AI agent with external data through RAG or other kind of services. Those are there to help us deliver solutions that help us as humans with our day-to-day business, with our operations, with our insights into data, etc. Delivering new capabilities into JD Edwards, right, AI, yes, what we're doing at ERP Suites is taking technology and augmenting the capabilities that the JD Edwards apps have today. And as we talked about yesterday, we're doing that in a way where we're not customizing the JD Edwards apps, which is something that we all know it comes with overhead and costs, but we found a way to do that without having not to customize your app. So you can still get the patches updates that you need from the Oracle, JD Edwards teams. And AI solutions for business aren't standalone products, right? They are integrated, you know, they typically on a cloud platform, the one that, you know, we've discussed quite a bit as ERP suites has been Oracles. And again, Oracles can run, AI can run with your JD Edwards. Whether you're running on a premise, whether you're running on a third-party cloud or you're running on a private cloud, OCI AI will work for you as well.


Why AI?

So why AI? And I'll go through these quickly because I want to have some time to show some use cases and demos real quick. Why AI, you know, increased investment in the last couple years has gone up by 300%, right, in AI methodology and we talked a lot a little bit about that yesterday regarding a methodology being necessary to make sure that you have a successful and fruitful AI adoption project. 57% of businesses expect AI to improve experience and support. And we're seeing that, right. Even in just the last 9 to 12 months, I've noticed the digital assistants or in some cases they're chat bots, right? Because a little bit more structured and, and rudimentary, but they're getting more sophisticated in the types of questions they asking, what kind of tasks they can complete. And, and by doing that, they're, they're doing it by learning, by learning in the interactions that they have with us. And that's what we'll show a little bit of today with the financials digital assistant, how it will learn, how we'll give you insights to your financials, particularly when you're doing financial planning. 80% of executives have indicated across a wide array of surveys that it improves productivity and creates new opportunities. The new opportunities we spoke about yesterday, which are in regards to new products, new services, new ways to message, new ways to do marketing as well. 63% of organizations agree AI is fostering innovation. That's what I just talked about. So just reinforce that point and consumers prefer interacting with digital assistants. Now that one's a qualified one, right? If, if it's a digital assistant that doesn't help, you're not going to like that, right? If it's very robotic and, and very rudimentary, you're not going to like about it. But if it saves you time, right, Having to wait, we've all had to wait on, on a, in a phone queue to get a human to respond our question. And then they bounce us around to three or 4-5 people and eventually we get our answer. If a digital assistant can do that and can do it with natural language and be conversant with us where we're almost seems like we're interacting with another human on the other end and we get our answer much quicker, Of course, we would prefer using that kind of technology.


Use Cases for Financial Operations

Use case, let's, let's talk a little bit about, you know, delivering insights, predicting analysis, productivity and automating tasks and minimizing effort. And this is a, this is a broad kind of swath of objectives that we put out here, but it's applicable to what we'll be discussing today in terms of the financial sector, right? Being able to drive better outcomes, right? Being able to do that analysis into the financial data and being able to get insights as to why 1/4 was so par. Or if you're seeing seasonalities that maybe have developed in the last couple years, but really haven't, you know, proactively managed that now you're seeing a slowdown in sales. It, it, it, it could be something that, you know, you can get in JD Edwards, or maybe you need a little bit more fidelity and be able to do that more frequently. While a digital assistant will be able to help you with that, be able to bring that, that data out of the JD Edwards database. And, and we'll show you a couple examples here in a few minutes. You know, the AI agent works 365 days a year, right? It's always up minus any, you know, maintenance windows that we all know are part and parcel of technologies, automating insights, analysis, reporting. The real interesting thing here is interacting, whether it's a digital assistant or other AI services where you can talk to it like you would just talk to, or maybe I would talk to Michael and ask Michael, I need your help with pulling this data out of JD Edwards and Michael then, you know, has to do some work. You can either run a report or you can figure out why this has particular query. I'm going to go ask an SQL expert to run an SQL. Those are the kinds of things now that you can do with natural language. And in that particular point, they'll be another keynote. I'll pitch a keynote that is slated for tomorrow by the Oracle Ask Data team where they're really making that literally translating natural language into SQL and then producing tangible insights from, from the JD Edwards data.


Franklin - The Digital Assistant

OK. And specifically as we know the JD Edwards team has talked about the JD Edwards data that you all as customers have built out over decades, years, etc. That's your digital gold, right? Well-being able to mine that through natural language is pretty pretty pretty. Sorry, question. OK. So serverless financial information integrated with JD Edwards apps, the digital assistant right in and we'll talk a little bit about how that's integrated, but and then and we'll demonstrate its capabilities at least a few of them with the time we have this morning. So I'd like to introduce Franklin. Some of you may have been some sessions yesterday where you all you've already been introduced Franklin. Franklin is our branding for our JD Edwards digital assistants. And we've, we've, we've built them for manufacturing, we've built them for sales, we've built them for other areas. Today we'll be focusing on the one for the Franklin for sales and it can perform tasks. It can create insights. It could analyze things for you. It simplifies querying data as well, grabbing financial data from JD Edwards. But it's more than just pulling data right and finding something in the FO 902 or in the FOFO 9/11 right, the general Ledger and aggregating it. It can also, it can also slice and dice the data and perform mathematical operations over it before it returns what you're looking for. And it's an agent that will be learning as it interacts with you on a continued basis. Enhanced decision making is what you'll get out of it, right? And, and reduce manual efforts by integrating a machine language, Oracle machine language. That's what OML is and generative AI, right? So there’s a lot being discussed today about generative AI. It seems like everything is being labeled a generative AI and it's important to know that really when you're creating something new or the AI functionality is creating something new, a new insight and something that is just not necessarily in there and displaying said something new. That's what generative AI is, right? And then you'll see some examples here today.


Demo of Franklin's Capabilities

So Franklin, I'm going to go ahead and run a demo here and I'm going to stop it in the spirit of time. I'll show a couple examples and I'll start the video here. I have things queued up, so we'll get this going and we'll move it up to a particular point. So what I'm going to set up now is total amount in the Ledger in. We'll start off with the very simple question. What's the total amount in the Ledger for Company 10, right? We'll ask it. So make sure I queue it up here in the right spot and you can see on the right hand side and hopefully you guys can see that. Oh, geez, sorry guys, let me go back to re queue that up. So here you'll see that and I'll shrink this down so we can see a little bit better at the bottom being time. What is the total amount in the Ledger for Company 10 and what it will produce here in a moment it will, it will acknowledge what you asked it and it will come back and say, OK,  it's zero. And you say, what's zero? Well, it's important to know what it's pulling. It's pulling, you know, the trial balance basically based thing is that's the overall balance. So it's basically by asking, OK, what is, you know, the question that I asked, it's basically asking it, OK, I want to find out if everything's in balance. And sure enough, it is right. So it confirms that.

So let's go with another one that's a little bit more interesting and show income now. So have it show income. I'm going to fast forward here a few steps and then you'll see show income by fiscal year where Ledger type is A. So we'll go ahead and advance that. It will acknowledge and it'll produce, you know, the, you know, show income by fiscal year for fiscal 15/16/17 and 18 and you can see in fiscal 15, right, Not a whole lot of income probably didn't know the company wasn't doing well. Who knows, it may have been a start-up. So let's, let's see where you know where that takes us. Let's grab some more information. The next example that I have is for fiscal 15.
Let's dig in deeper on that fiscal year. Let's see if we can get a little bit more insights and understand why that negative happened. So let's show fiscal year 15, quarter one, quarter two, quarter three and quarter 4 income. OK, so I'll skip to that section. Again, I'm scrolling through this so that we can make the most of the time that we have. And you know, we're going to try to get in some introspection on that fiscal year and see if we can get some more insights as to what might have happened and see if we are any seasonalities, maybe they're impacting us and so forth. So you can see now that it, you know, it produces the income by quarter by quarter different quarters.
The other thing that it has done is if you know in JD Edwards we don't necessarily have a defined table, not even the fiscal in the fiscal years definition the quarters. So Franklin here has figured out the quarters, it has pulled up the balances and you can see right that the profit here that has shown you can see some variations here and you can see that the numbers are negative, right? But they're negative because, you know, that's how they're displayed, that's how they're stored in JD Edwards.

Let's do one more example, then we'll continue our discussion. Let's look at, let's dig into Q2 and look at year over year of Q2 and see how that particular year's doing and they'll give us some additional insights as to what's going on. So let's forward to that. Simple one second finger, right?
Oh, you can't see the full screen. Let's see. Are you still seeing notes? Let's see. Hold on, let me stop sharing. See if I can share again. Let's see. And is this any better? Someone answer on the chat, please. That looks better now. OK, great. So we'll proceed with that. I'll see if I can get one more example in here quickly. We'll look at the amount of net sales by fiscal year, year over year. So we'll get back to the session here, see if I can queue it up in the right spot. So we're going to get the amount of net sales by fiscal year. It breaks it down. You can, you can see the, the negative amounts that are too, but that's again, how, how you know sales is, is displaying in JD Edwards 2015. Again, we looked at in the not a whole lot of revenue there. But then we, we, we're seeing that there's growth, right. So maybe, you know, some of our suspicions made of this a start-up year, maybe just sales was, you know, wasn't good in that year or you're seeing trends here.
So a lot of the things here, you know, and, and since we're short on time, I'm going to skip to our next slides. Hopefully this gives you a taste of the things that you can inquire. There's a lot more things that you can ask for. You can ask for profitability, you can ask for internal, for internal R&D numbers. It can handle machine ledgers as well. It's pretty, pretty flexible what we've done and it can interpret language. So you don't have to tell it the exact terms of what it does. It could be lingo that we use in JD Edwards that is, or it would use general accounting language. It will translate it into JD Edwards lingo to figure out the right kind of data. So it's pretty impressive here. This is the internal R&D example.


Conclusion and Next Steps

So because we're a little short on time, I'm going to go ahead and pause the video and move forward to the next steps. If people want to see more of this video or want a deeper demo, please don't hesitate to reach out. OK. So I'm going to go ahead and move forward. In the spirit of, of wrapping up here quickly, you know, Franklin's powered by orchestrations in the behind the scenes and of course Oracle AI we're using, we're using Oracle data warehouse for a couple of scenarios where using digital assistant machine learning, the general artificial intelligence offerings within, within the OCI collection of services as well as enterprise one. The key thing here is starting with 9.2 A2 JD Edwards enabled us to be able to directly authenticate OCI AI services within orchestrations. And that's what's made these possibilities a lot easier to create. So bridges interaction between, you know, user needs and, and, and the database system simplifies interaction. You can also see it's, it's a nice user interface, right. So Franklin is sitting side by side with the base JD Edwards application. So you can see the context. You can, yeah. And in some use cases you'll see that rules will be highlighted. You'll see things being able to be pushed back and forth. If that's why, if those are the scenarios that you need, it's pretty flexible. So we built this so that again, right, best user experience possible. I talked about the machine ledgers automating a PNA task as I talked about earlier is a key thing. We talked about efficiencies quite a bit with AI and this is no exception..

Cloud infrastructure, I'll skip to do this quickly. We're part of the RAG beta and gotten a lot of insights where we can, you know, augment the knowledge that AI has with external data, but not let your data leak out, right. So it's secure. Some additional illustrative examples that unfortunately we will run out of time to show this, but you can see how the other kind of questions that you can ask Franklin. And in conclusion, you know, embrace the future, which is AI. It's the present and it's going to continue to evolve now create new capabilities and then embrace the productivity gains that it comes with. So thank you for your time. Apologies for some of the technical challenges we had today, but hopefully this give you a little bit of taste of what's possible. If you have any questions, please do not hesitate to reach out to me and we certainly could have a follow-up conversation on any topics or show you more of what Franklin is capable of. We've talked about the AI journey in the past and of course, we'll be a blueprint. We'll be a user groups and keep an eye out for our video podcast that we have that Nate leads from our team. It's a fantastic one. Enjoy the rest of the day and we'll be in touch.

 

 ChatGPT

Nate Bushfield

Video Strategist at ERP Suites