This session highlights the ongoing evolution of JD Edwards, Oracle’s long-term roadmap through 2036, and its role as a platform for digital transformation. The session covers key themes including customer adoption of hybrid cloud and OCI, the growth of Orchestrator, and new innovations such as the Enterprise Process Modeler and EnterpriseOne Widgets. Presenters emphasize Oracle’s enablement strategy for AI, detailing integration with OCI AI Services like document understanding, vision, and generative AI, alongside practical use cases such as invoice processing, order capture, and chatbot-driven insights. The event concludes with Q&A, resources on LearnJDE, and reflections on JD Edwards’ nearly five decades of innovation, offering organizations a clear view of how to harness automation, cloud, and AI within the JD Edwards ecosystem.
Table of Contents
Transcript
Welcome & Introduction
Again, welcome everybody to day two. I hope you enjoyed yesterday and we've got a full schedule ahead today. So, stick around and I'll turn it over to John.
Thank you, Scott. Do we want to wait any longer? We'll just get going. I see a few more people trickle in, but uh welcome all to day two of ERP Suites AI virtual week. So, we're excited to be here.
My name is John Green. I lead the JD Edwards product management organization at Oracle. I'm uh joined by a few of my colleagues AJ and Brashant. I'll let them introduce themselves when they get to their sections, but we're excited to be part of this event. Um you know, a first time for ERP Suites. So, thank ERP Suites for putting this together. Um we're definitely seeing momentum in the market and customers. It's a top of-mind topic right now, right? AI.
Um, as we were putting this road map presentation together, we were looking back at old road maps and even various conferences that we've done live and virtual. And it's interesting to look back 2018 or so. We started an intelligent ERP. What is JD Edwards? Can it be an intelligent ERP? What would that look like back in 2018? Here we are in 2025 and these start things are starting to come into fruition, right? You can see the capability of AI, the investments that Oracle is making in AI and JD being able to take advantage of that.
Orchestrator History and the Future
So it's fun to look back 9.2 release in the market in 2015. Our orchestrator product that we delivered around that time, AJ's baby as we call it baby or in 2016 or so was the first thing it was called was the internet of things orchestrator right it was the intent of that in that time and look how far it's evolved and now it is an enabler for taking advantage of AI services from Oracle or other AI services out there and any of the other OCI services available from Oracle as well not just AI so we're excited to be a part of this conference to share share what we're doing and our road map overall.
So, it's not just going to be an AI focus even though this is an AI conference, but we will dive deeper into uh AI at the end of this presentation since that's the focus of this week. So, we're excited to be here. Thank you all for attending and let's get this going.
So, as I was talking about the looking back 10 years, we also like to look ahead 10 years. I mentioned 9.2 was released in 2015. The original end of premier support date for 9.2 was supposed to be 2023 but here we are in 2025. It got extended to 2025. Then it got extended to 2028. It has been if you had not heard in the March April time frame just recently was announced that we extended that through 2036. So it's called a rolling 10-year window. I'm not an expert in math but it's actually 11 years right now. Right. So we have 11 years to look ahead of premier support.
I like to call it a not even it's as a support timeline it's a development timeline. So it's providing new features and functionality to the enterprise 9.2 code line. So it's reassurance that we're going to continue to invest and deliver new features and functionality to the market.
Why was I talking about 10 years in the past? We're also looking at what does JD Edwards look like in 2036? What is the next wave of innovation? What are the new emerging technologies that we need to make sure JD Edwards can help our customers capitalize on? We didn't know exactly what AI is going to be as it's evolving now and it's still very new still a lot of questions covers like this for customers to start asking questions figuring out use cases how would it apply what are the the benefits um AI is not always the solution right it is another tool in the tool bag to solve business problems but our road map as we look out 10 11 years is whether it's AI or any other emerging technologies, how do we make sure that JD is ready to meet your business needs and solve their business problems?
JD Edwards Roadmap and Long-Term Support Strategy
So, we're excited about the continuous delivery uh in 2036. It is for all of applications unlimited customers. So, it's not just for JD Edwards if you're not aware of that. So, the major product lines that make up applications unlimited are EBS, Peopleoft, Seiull, Hyperion, and US. So, five major product lines, thousands of thousands of customers, and like I said, this is the 10th time it's moved. I don't see that changing in the foreseeable future. So, 2036 is our newest date. Excited about that.
When we uh do various events like this, whether it's virtual or in person, we like to talk about how your success and how our success together. It's great to see everything that's been going on in the product.
So, for Enterprise 19.2, two. Uh we've had 30 plus either apps or tools releases since the GA back in 2015. Just in the last uh year alone, we've added over 110 new features across applications and tools. Um looking all the way back since 2015, since the GA day, over 950 new enhancements in the software.
We also like to track our customers in their adoption of cloud. So over 470 JD ever customers that are what doing what we call hybrid clouds. That means they're running JD Edwards at the core and they're surrounding with fusion SAS applications like uh OTM, CPQ, EPM, um HCM, various uh fusion SAS offerings complementing their JD Edwards at the core. So that's what we call hybrid cloud. So it's great to see so many customers doing that.
Then another thing we track is JD customers that are moving their environments and running them in Oracle cloud infrastructure OCI. Um this is obviously Oracle's public cloud offering. Yes, our customers can run in Azure AWS Google Cloud Platform in OCI.
We give customers choice and control but we obviously track our JDR customers that are running in OCI to take advantage of all the benefits of it's not just a cloud environment. It's the services like AI services that are in OCI that our customers can take advantage of. So 3450 customers taking that on.
Uh the baby JDS orchestrator over a thousand customers using JDRs orchestrator at this point. I actually like to say we should count this the opposite direction. Instead of counting up of how many customers are using it, we should count down and say who is not using orchestrator at this point. Right?
If you attend any uh uh conferences, whether it's the Quest virtual conferences or our live blueprint 4D or in focus, there's always tons of sessions where we see great use cases and our customers solving complex problems uh using orchestrator. So great to see so much adoption in the market. Like I said, that came along with 9.2. to many, many years ago. It has been wild to see how it's evolved and it's one of our key pieces to help us take advantage of AI.
I said it was with my baby. I didn't mean to say I didn't love it, right? But, uh, I'm sort of like an uncle. Let's say I'm like an uncle. Oh, orchestrator. Yeah. My my godne or maybe maybe it's a she. Is it like a ship? Maybe orchestrator is a she.
Yes. My wife is the orchestrator on my house. So that I think that's appropriate.
All right. So uh the JDRs ecosystem over 150,000 learn JDE visits in the last 12 months. And I uh we'll talk about this at the end as well. The JDS connection podcast. If you're not familiar with that, it is a new podcast that was launched about a year ago. We've had over 9,200 uh downloads and listens to that podcast. Like I said, I'll talk about that later on. So, if you're not aware, I have some more information on that. Or if you are aware, thanks for being one of those 9,200 uh downloads and listens.
Customer Adoption and Hybrid Cloud
Something very near and dear to my heart is go lives and successes over a thousand plus go lives or upgrades. This we started really tracking around the uh start of COVID because we did see some customers having to delay some projects or postpone, but then we saw some customers accelerate projects at that time because they needed to get some things done.
Uh and so we really started tracking go lives. Our support organization tracks these. Um there's a process for you as a customer to log a SR to get that milestone logged and then the support team shares these with development leaders, product management leaders and we review the customers and that's how we know hey look at all these customers that are going live already on release 25. Um, it's great to see that. And then with those go lives, we turn that into over 450 plus customer success stories. So hopefully, uh, we'll see more and more of your go lives being tracked. And it's great to see that number continue to climb as customers move forward.
So like to dive a little bit deeper into our overall product strategy. If you're not aware, three major components of our strategy, applications, digital platform, and system administration.
So for applications um you know we have a long history of developing and delivering applications in the marketplace all the way back from our world days. Um and one of the podcasts recently we were talking about the history of JD Edwards and we actually started in 1977.
Um that's ironically the same year coincidentally the same year that Oracle started. So 1977 we started developing applications. So, so much intellectual property, so much uh code and and functionality in our applications, but we're always trying to continue to simplify the experience. How do we make it simpler to use the applications? How do we add features to help automate business processes? And continuing to enrich feature functionality. Our SIGs do a great job of saying, "Hey, I would like to see this to be improved to enrich that feature functionality." So adding those enhancements to the product specific areas to benefit our customer base.
The digital platform um this is where the orchestrator falls. This is uh uh AJ's area of expertise. So using the digital platform for process automation. It's our low code no code framework also for enterprise integrations. So much you can do with a digital platform. What I like to highlight here is when we make enhancements in the digital platform, these benefit any and all customers, right? When we do something in applications, it will maybe benefit customers that were using that specific application. Uh but not every customer use their applications the same way. When we do something in the digital platform, it's in tools uh updates and it benefits any and all customers. So you'll see a lot of uh feature functionality being added to digital platform over the years.
System administration. Um this is where uh Pashant falls and a lot of his work is and so it's making sure that JD Edwards can run on all the modern infrastructures public clouds or on premise we give you the choice of control where you run JD Edwards and how you run JD Edwards uh sentinel administration making sure that it is easy to administer lowering the total cost of ownership of running JDS and predictive patching how do we make sure that you're applying patches that you know what you're uh applying thing, you know the impacts of that, you know where to test for that. So, constantly trying to lower the to total cost of ownership and making it making it easier to manage and maintain your JD hours.
So, hopefully you're aware that uh in the October uh November time frame, we released uh released 25 uh into the marketplace. um a few years ago back in release 22 we moved to this uh naming convention of 22 23 24 25 and that's simply because you know as AJ and Pashant and I talk to customers and I say I'm running 924.3 and I'm sitting there and I have to do some I have to put it into AI right and just say what year was 9243 to figure that out but now when a customer says hey John I'm on release 24 I know you're at about 2020 24.
So, I mentioned we have premier support through 2036. Guess what the release name will be in 2036. For all those playing at home, it's release 36. Exactly. So, see how easy that is. But I'm still a a diehard and under the covers. Release 25 is 929 and update 9. But anyways, you get the idea. That's why we have that releasing naming convention out in the marketplace. So, makes it easier for all of us to track that.
So, release 25, as I said, we've seen some strong adoption since it was released in November. Um, it's great to see customers taking that on. I have a number of customers. I always call it a race. We love to see the first customers that go live on that. And we've had plenty since we're here in April. Is it still April? April 2025. And we have a number of customers that have already gone live successfully on release 25.
So uh I'm not going to go through every new feature uh enhancement in release 25 up to this point but we do want to give a few highlights. So simplified user uh experience for applications. We made some enhancements in license plate workbench uh usability. Uh this is a financials enhancement ability to assign delinquency fees to the business unit of invoice. And this last one here is a real estate uh enhancement. So ability to edit tax information and manual billing lines.
Automate business processes. So you're going to see all four of these are under the category of enterprise auto automation and that is a topic that AJ is going to dive deeper into today. So I won't dive into that right now. We have a number of E1 pages in our business process model that AJ is going to talk about and maybe even demo.
So then the enrich feature functionality um we launched our first phases of uh the sustainability framework. So this is taking advantage of or looking at uh ESG environmental social and governments. We have our first code against uh the sustainability framework. Two-way voucher match automation was delivered with release 25. We actually delivered three-way voucher match automation a few years ago. Circle back now and with release 25 we have two-way voucher match automation.
Increased range of account ID. Um, we had a number of customers that were running out of the account IDs because it was a numeric field. So with release 25, we've enhanced that. Now it can be an alpha numeric field. So you can have an unlimited number of account IDs. So if you're running into that, take release 25.
Product Strategy Overview
On the digital platform side, we're gonna have AJ because this is he's the uncle of all of this. So go ahead and I'll let you talk to this.
Yeah, I mean as far as roadmap goes um lots going on in the digital platform as John said this is the the magic layer that when we do something here every application every form every experience benefits from it integrations and um process simplification automation uh the extensibility digital platform includes the extensibility framework orchestrator workflow um the foundations the infrastructure for all we have going on with enterprise automation.
So you know when your boss comes to you and says hey um AJ for the for the overview for the road map can you give me one slide that shows all the road mapap stuff for digital platform and that's like yeah can you put the ocean in a cup in a in this cup for me. So, so much going on, but here are a few of the highlights and uh it's also interesting because release 25 came out. So, most of this roadmap stuff has already come to fruition and we're going to uh take a look at this.
Sure. John, put me on the spot. You're going to make me do some demos, John. All right. Uh sure. Throw throw that out there. But um yeah, we got this the probably the the biggest news in the enterprise automation space is we rolled out the enterprise process modeler and we talking about enterprise automation as a initiative for about two years now but and we always had this enterprise process modeler thing on our our vision and you know vision became strategy and strategy became roadmap and roadmap became product. So the enterprise process modeler is now generally available with release 25.
That's pretty cool. Uh just hot off the press when I say this roadmap is coming to fruition. Enterprise one widgets. I'm pretty excited about this. Um and in a nutshell, we all love watch lists, don't we? Well, what if watch list could sort of take on a an entirely new um level of of uh ease, efficiency, and capability?
So we'll do some demos on on widgets again generally available. So this this isn't even road mapap anymore. We can see this uh kind of coming back in the not in the in the dashboard in the in the windshield but now it's in the rearview mirror. And then we're also doing a bunch you know to the enterprise one pages to make it line up and and be more effective in the context of enterprise automation. Right?
So we know we can build enterprise one composed pages quickly and easily. Um and we we jumped on that to help us build process models um sort of as a as a um as a pallet as as a way to drop specific uh um nodes if you will, right? steps in our business processes and then connect them with arrows and then superimpose metrics on top of those in the form of watch lists. So if you've been watching closely the uh enterprise one pages have been picking up more and more functionality bit by bit in an agile fashion and of course we want to do all this in a no code low code fashion.
We want to empower uh business analysts who know the business very well, who know the problems, who have a passion for serving the business, but just might not happen to have coding skills in a particular language. You don't need to know JavaScript. You don't need to know Java. You don't need to know Groovy or any programming language per se.
And we also want to be able to develop these um enhancements, right, that in a in an agile fashion. We don't have to want to have to go into form design aid and crack open development tools. We don't want to have to build packages and deploy them out. Oh, I'm sorry. We only we only uh build and deploy packages on every other Tuesday of the month, right? Business happens faster than that. So the the no code low code approach with the user defined object framework um you know that's what we're after to be able to invent solve problems within hours or a day or two of course all within good governance right all within um you know promotion and and testing because these things affect the business but you know we want to be able to do these things in in days or so not a six-month development project.
Um and of course you know as uh the strategy statement mentioned earlier John showed that the digital platform strategy statement we still realize as a matter of fact now more than ever that JD Edwards your enterprise one system is not on an island. We we need enterprise integrations.
There's so much going on in the digital economy right now and most of that most of that comes under the umbrella of what we would call cloud services right there's all kinds of third parties to integrate to and we've been doing that for 30 years right no surprise and if you look you know again we'll take John's theme of let's look back 10 years 20 years historically we've always had the need to integrate Right?
Was it the comm connector? Anybody remember the Cororba connector? Right. And we used to do hard brittle pointto-point integrations that way. And then um you know we got to Z files and you know XML exchanging XML files and there's still good place for that. It's efficient and you know it's it's solid. Um, but it's technical and it's complex and if you want to do one of those EDI integrations, you know, it can be a bit of a a technology project.
Uh, then we got to business services. That was a big step forward, right? The the ability to describe business objects. Um, but again, you needed some deep deep understanding of the JD Edwards business functions and to wrap those business functions in business services. and then have Java skills to put Java wrappers around them and build business services packages and deploy those and have a you know the complete other business services server. Certainly, it was a step forward, but the the world's gotten better in enterprise integrations.
And you know, it's mostly because of REST APIs. And now with Orchestrator being able to call outbound to REST APIs, that really kicks the door open to JD Edwards participating in the digital economy in a no code, low code fashion. And you know, I'm, you know, what I like to say is I'm not a uh I'm not a developer, although I did get a B in Forran in 1981. Um, but I I have no skills in coding right now, but I can see problems and I can see solutions to problems and I spend some time training myself up on the tools and it is absolutely possible for me to build orchestrations that go out to thirdparty REST APIs, supply them with some input.
Nowadays we like to think of that input as a prompt. have that thirdparty service do something for me and get a response back and then taking that response and being able to do something smart with it in JD Edwards, right? Is a pretty simple pattern, but when we look at that simple pattern, um almost like putting two hydrogen atoms and one oxygen atom together. Um that's a pretty simple pattern, too. But now we've got water and and there's lots of things we could do with water.
So you know this enterprise integrations thing shouldn't go um um underestimated especially and I think pashant is going to talk more about this right the orchestrator authentication to OCI services and all of the um cloud services that are offered by the Oracle um OCI um list um and now that we can authenticate to those and we can orchestrate those um opens up just a huge huge possibility abilities.
So, that's my five or six minute spiel on the digital platform. Um, I could do this for a day, a couple of weeks. We we could have workshops on this, but there's your one slide, John. There's my my one slide feel on what's going on with digital platform. And as a matter of fact, like I said, most of this stuff is not roadmap anymore. It's all come to fruition. So, that means it's time for a new slide and a new road map and, you know, new a year or two full of full of new stuff. So, um I I I can't I can't uh I can't spill the beans, but our team is working on that right now. Right. I can say that. Yeah.
Yes. Exactly. Lots coming in release 26 and beyond. Right. We're going to do it again. We're going to fill up another slide and make it come to fruition. So, that's right.
Orchestrator, Poll Results & System Administration Roadmap
So, thanks, Uncle AJ. And I'm going to stop and look at the the poll here. So, Scott did throw a poll out there. How is your organization's use of the orchestrator?
The orchestrator. I like that. And it's a majority use it often and can't get enough, which is what we like to hear. But Uncle AJ, you actually have a few responses saying you have an ugly baby that they don't use it at all. So look at that. They just don't even like it.
No. Uh we understand uh there's some customers out there, like I said, the countdown. We're going to put that. So we have a few customers not even using it yet but hopefully uh you know there's tons of information out there and learn JDE obviously a good rundown of all the power of the digital platform not just orchestrator um but take advantage of that I think you'll see it's a major building block for using AI right so and here's a quiz out to the audience July we got July coming up right it's a tomorrow is May and we got June and July what is the significance of July July for orchestrator.
Pregnant pause 10 years in in July.
Yeah. Oh yeah, you can do that. Yeah. Great. Yeah. No, in July orchestrator will have been in the market for 10 years and it still feels new, right? Still, like I say, 21% of, you know, the polls here, people not using it. Still lots of opportunity. It's there when you're ready. Um it's as easy or as complex as you want to get into it. So um lots of good um uh training out there, documentation. Come to Blueprint. I think I counted what did I count 20 21 orchestrator sessions in in the four days at Blueprint. So yeah.
Nice. So you have a 10-year-old basically coming up. You're about to have a 10-year-old. Imagine that.
Yeah. I'm I'm looking forward to the days where orchestrator becomes an angsty teen.
Uhhuh. I was I say and then uh can vote when they're 18 and stuff. I mean imagine what orchestrator can do when it has that autonomy and independence like my 18-year-old. But anyway, let's not get too much into that.
All right. So, uh rounding out the last area, the focus area is system administration. So, the road map for release 25 sentinel administration simplifying the dev client. So a simplified uh web tier. So really making it uh easier to uh administer uh the centralized web tier. Uh that came out with release 25 seen a number of customers take that on.
Uh UDO userdefined object altering these went gang busters. Our customers loved using udos. But then a good sig enhancement request was hey we have all these udos but we need to change the titles or we need to reclassify. We need to you know change you know whatever we could with them. the the altering we added that with release 25 so you can do that now this is a cool one here as well optimize package full build in our internal testing we were able to reduce the full package build times by 80% so trying to see what uh the impact and benefit will be for customers how much they can reduce their full package builds so kudos to the dev team on that uh feature modern infrastructure server uh manager added developer access directly in server manager also performance workbench direct access uh an integration to server manager.
So for those that use performer workbench on that for oneclick provisioning for anybody that does provision using our oneclick either onrem or an OCI now we uh have open tofu support for our scripting language there and predictive hatching uh custom multi-platform deliverable support. Then the web update reset merges. This is a popular one that came from a lot of our partners and customers alike that basically as they do the the merges being able to reset that quickly to be able to redo it. And then the last one's web planner ESU.
So lots of like I said for system administration lowering the total cost of ownership making it easier for CNC's to uh run and maintain JDRs but how do we also make it where you can automate a number of these processes right so we've been webifying a lot of the uh functions in these processes so that way they can be automated so lots more work to be done in all these areas but that's just some sneak peek in what we've done in release 25 so far.
So release 25 and beyond. This is our acronym slide. So we have sustainability framework, enterprise automation and AI. So this is where these are kind of the three of the main initiatives we've been working on in addition to our overall product strategy.
Three major areas of investment that we've been doing in JDRs at this point in time. So ESG uh environmental, social, and governments. We're focused on the environmental aspects of this. We want to help you uh simplify and streamline the environmental data collection process whether it is due to legislative reporting needs you know by a country or by specific industries you know many countries especially in Europe Canada uh it's becoming a legislative requirement California the country of California as well but uh whether it's for legislative requirements or c customers uh that simply want to support their sustainability goals on their websites and show here's how we're doing the best that we can for being sustainable.
We added a whole new system code 20s and we added a new uh sustainability activity ledger in release 25 to help you capture in various applications right now voucher match and standard uh vouchers that you can add this environmental data to the sustainability activity ledger.
We're doing this as a data store, right? We're helping you collect that data. If you have this currently being done in spreadsheets, you can import the spreadsheet to the sustainability activity ledger, but then we're leaving the environmental reporting choice and control for you to choose what solution you would like to use. Fusion uh SAS has EPM that they have a specific reporting solution for ESG. So, we've piggybacked on that and we say customers can use that. We actually published a white paper on how to integrate JDRs with EPM to do that. But if a customer has a different reporting solution for ESG, they can use that as well. We're just basically helping you capture and store the data in JD Edwards so that then you can report and uh calculate and do the things you would like to do in whatever thirdparty system you choose.
So that's a high level overview of sustainability framework. We do have a long road map. we've done so far just scope one, but we'll be starting to deliver some of the scope two requirements and looking at scope three in the future and beyond.
Enterprise Automation Framework
This is where I'll get to pass it back to Uncle AJ to talk about enterprise automation, another one of our major initiatives.
Yeah, thanks. Thanks, John. And, you know, for once, I mean, I guess I already broke the rule because I wasn't going to talk too much about orchestrator. Um, but uh we call that the cowbell. I don't know you guys. um seen the Saturday Night Live skit with Will Frell and the the cowbell in the rock and roll band and you just can't get enough of it. But we're going to pause. We're not going to talk about orchestrator a whole lot. U we're going to rotate into enterprise automation and this is an another initiative, another way of looking at the benefit that we can ek out of our JD Edwards system.
Um we started this about two years ago and again you know looking back a couple of years ago it was um it was opaque. It was opportunity for creativity. Uh we had some vision and that vision turned into a product strategy and that product strategy turned into actual features on the road map and we've been u making steady steady progress to those features.
But the idea goes like this, right? You know that in your JD Edwards system, you've been collecting data transaction by transaction for a long time.
And the reason we do that, the reason we have an ERP system, the reason we take the trouble to record those transactions, whether they're work orders, sales orders, journal entries, leases, you know, whatever it is, the only reason we collect those and store them in a database is because we have a a hope um and an optimism, a dream that that data one day is going to be useful.
And transaction by transaction, we've put steps, you know, foot in it that carve a foot path in the grass. And now we can take a step back and and look at that that footpath in the grass and look at that data.
And we have a hunch that that data which now represents your your enterprise, right? that data, if you take a look at it from a high level, really is a digital twin of your business, right?
You you can create powerpoints, you can create VO diagrams, you can create um statements on on your company's website of what your mission and your goals and your objectives are. But if you want the truth about how your enterprise really operates and what what you're doing, you should step back and look at that data.
Some of that data is going to be obvious to you. Oh yeah, I I know how my business operates. But we have a hunch that there could be blind spots. There could be areas in that data that says, you know, yeah, I knew this happened. I had no idea it happened so often, right? And these are the opportunities for improvement.
Whether it's a bottleneck or you know a a slow spot or a problem in your your business operations or on the positive side it could simply be the opportunity for a new business model. Hey, I didn't know I had a gap here. Boy, if I can attack that gap, maybe I can sell new things. I can sell more things. I can sell things a different way. Uh so those are the blind spots that we're we're looking to um to find in in this whole framework that we're calling enterprise automation.
Move forward and as I said right yeah we can move forward. There we go. Thanks.
Um, you know, as I said in in ERP land, in your JD Edwards system, we attack that transaction by transaction. I don't think this screen is going to surprise anybody, right? We get we enter a purchase order and I know that I can enter purchase order 1 2 3 4 on on a certain date and that's recorded in the system as a transaction. With enterprise automation, we're looking, you know, at a at a much wider, broader aggregation of that. Hey, there's 144 purchase orders created this month.
And and I know for many customers out there, that might be more like 14,400 uh purchase orders created this month, right? You know, the volume of data that that you're generating. And what we're looking at is to slice it and dice it and say, well, hey, I know that I have a process, a business process. My uh let's say my order activity rules say that a purchase order needs approval and some of those approvals get rejected. I I have a process for that.
Oh, but I didn't know that, you know, 55 of them got rejected this month. And out of those 55, I didn't know that 51 of them were rejected by the same supplier. You know what? Gee, that maybe that's telling a story. Maybe that's a blind spot. Further to that, I didn't know that 50 of them always get rejected on Fridays. That's an interesting clue, isn't it? What's going on?
And you know, as we dig into this data, you know, we can find you and reveal these blind spots. But it's only when we take this data out of our JD Edward system, we aggregate it, we slice it and dice it because because let's be honest, we've been doing that for years and we call it business intelligence or maybe we call it analytics, right? We we've had different names for it and we know we want to pull this data out and look at it in different dimensions.
But in enterprise automation, the important piece is to superimpose those analytics in the business, right? with the business process, right? We want models and metrics together because it's only when we can see those metrics and the analytics in the context of the process that we can figure out how to change the numbers. Right? If you want to change the numbers, you have to change the process.
And that's where the, you know, on the left hand side, that's where the innovation comes from. that's where, you know, we're going to be able to apply human ingenuity and solve those problems. So, yeah, John, we can move forward.
Um, so this is this is the framework we use to to visualize enterprise automation. This five-step circular life cycle, ingest, model, analyze, solve, and measure. And we break it down this way.
Ingesting means we need good, clean data. And as I said, many of us have been collecting data in our JD Edward systems for years, possibly decades, you know, and so just think of the pile of digital gold that's in our databases.
And all of that data is is available to ingest. And by the way, this is this is the intersection with automation with process automation because the more we can automate the collection of that data, the more we can take people out of the the business of collecting data manually, right? People make mistakes, people are slow, people get tired. Um, if we can automate the collection of that data, we have not only more digital gold in our database, but that digital gold is of a higher carrot value. It's not 10 karat. It's not 14 karat. Right?
With good automation, we're getting 24 karat gold in in our databases. And so now with all that data, we have an opportunity to build tools to read into our JD Edwards data and model that um model it automatically. And I'm going to show you uh that enterprise process modeler that we built and uh a modeling technique with enterprise one pages.
So we know we have tools to do that and and also to pull out analytics out of the data and put those in the context of models, models and metrics. That's where the blind spots start to appear. That's where we get some good vision into what's actually happening and we get to put our good old human ingenuity into solving those problems. We're in the solve phase and then we're proud of ourselves, right? We we built more automation, maybe we built some orchestrations.
Um, maybe we did better um, negotiations with our trading partners. Maybe we adopted some JD Edwards functionality like three-way voucher match or something that's like right in front of us in our JD Edwards system that it's available to us and we're just not using, right? There's lots of ways to solve the problem. And then we're proud of ourselves and we sit back and say, "Boy, that was a great project. I did a no code solution to solve the problem. It took me three or four days. I rolled it out in a week and everything's better.
Or is it? Well, just like, you know, back in the ISO 9000 days, you can't improve what you can't measure. And so, we need to be measuring that and capturing new metrics and being able to see, you know, did we did we actually take a step forward? Did we improve? And the cycle repeats. So, that's the framework that we're using for enterprise automation, and that's what we're after.
Modeling, Metrics, and Analytics
But the core of it really is, and and I'm going to slip into some product demos here in a second. Um, but the core of it really is models and metrics. And so, um, you see in this slide, I've flattened out the ingest, model, analyze, solve, and measure steps. And we're not going to go through, um, all of, you know, the five steps. If if you want to though, just tap me on the shoulder because I'm ready, willing, and able to dive deep into all of the any all these five, right? But as I said, the core of it, the really interesting part where the rubber meets the road is modeling and metrics. And we have a couple of ways to do that. And the, you know, the couple of ways each have their own unique charms, benefits, drawbacks, advantages.
And the first one is sitting right in front of us with using Enterprise One pages as a pallet for um uh for drawing process models. And many many customers have done this already. You know, talking to to customers in um in webinars and at conferences and in through email and chat. Oh yeah, we AJ, we already drop tiles on an enterprise one page and we already connect those tiles using arrows. And so that really is a quasi process.
What had been missing from from this? Okay, now I have a model. I I I can use Enterprise One Pages to build a process model. What I'm missing are the metrics. And so we did that feature. Hopefully you're aware of it. if not, you know, look at our product catalog and our announcements. But a pretty simple feature to put watch lists on the tiles.
Well, now we have now we have metrics that are in the context of the process model. And that that gets pretty interesting, right? Once we can drop these these watch lists on the edges there. And that's what you're seeing with those those blue um those blue ovals. Um there's a couple other things missing. that uh we've built out and I'll I'll show you that in the demos.
So again also the analytics right a lot of you know of our UX1 role-based analytic pages and we've rolled out a whole bunch of uh charts happen to be based on jet um that adorn that you know go with our alert analyze act model right with UX1 we have alert analyze act and with enterprise automation we have ingest model, analyze, solve, and measure. And it's no accident that analyze is the middle piece in both of those.
So those UX1 role-based analytics and the the charts that you see there, I think the next build, John, uh, click again. Right. That those are all completely reusable and interesting in the context of enterprise automation.
But um the benefit now is that we get to see them in the context of a process model and and yeah I mean once we start um rolling out these tiny little enhancements it's like who let the dogs out right let's go let's build some process models let's get some analytics going there. So um right in front of our face without doing any you know major um enhancements we can use enterprise one pages.
Now, let's take a look. If if I'm going to build those process models manually by building pages, well, that's not much better than what we used to do, you know, back 10, 20 years ago when we sat in a conference room in front of a whiteboard and we we decided what our process model was going to be, our enterprise processes, and um we we put them in Vizio diagrams and they were sort of set in stone and they were the the hopes and dreams of what we're going to be.
What if in that ingest phase we could ingest some JD Edwards data from somewhere um and automatically draw the process model not based on what we hope it is not based on what we we think we want the design to be but as a reflection of what our enterprise data actually says right what you know what is it what is it that um we could be we doing with with that digital gold and so that's what the new enterprise process modeler does.
And hey, order activity rules is a great example of data that's in our JD Edwards system. And you know, for those of you that know and use and possibly have built order activity rules, it's we know that purchase orders and sales orders and document types have certain statuses and the order activity rules define what status is allowed to go to what next status.
And you can imagine that those sets of rules that in that little red box there, right, in the JD Edwards system, it's all in table format. But wouldn't it be more interesting if we could take those rules and visualize that as a dynamic process model? And that's what we've done with the enterprise process modeler. We'll take a look at that in in real live demo.
So the next is yeah, a view of that process modeler. You'll you'll get copies of these slides, I think. Um there's plenty of them out there in the public domain. But not only do we have the automatically generated process model, we have metrics on the nodes and links between the nodes and we have analytics over on the right hand side that are in the context of you know what those nodes and links can do.
So we're pretty excited about that. Like I said, this process modeler is generally available. release 25 as of last was it October we released it in November. So um um yep that's that's out there for you now.
I think this is Oh yeah, one more topic. One more topic and then we get to see it live. Yeah, watch lists. We all love watch lists, don't we? Don't we? Pretty pretty simple way of getting metrics out of the JD Edward system. I I love watch lists because of their simplicity, right? because of um yeah I go to an application I can set up a query in a couple of quick clicks I don't need to know any SQL right we're right back to that no code I don't need to know SQL but uh if you knew it or not when you're creating a query in your enterprise one application and filtering the data that's in the grid you're actually slinging some SQL statements behind the scene and that's the no code beauty of watch list and then with two more clicks I can take that query and turn it into a watch list.
But the drawback is that watch lists always come from a query and the query always comes from a row count and that row count always comes from data that's in a grid. And at phase value, that's still very very valuable. There's nothing wrong with that, but it sure is limiting.
What if what if the number in a watch list could come from an orchestration, not just a query with a row count. What if and you know for those of us that have been building have been building orchestrations, we know how powerful what orchestrations can do with form requests and data requests and all. So now that the number that an orchestration can generate doesn't even just have to be an integer of a row count of a grid. It could be calculated. It could be manipulated. It could be aggregated. It could be an average. It could be a sum.
Um so that gets pretty exciting. Well, what if the orchestration generated content that wasn't even a number, right? What if I want to see, hey, what is the my most ordered item name? Oh, well, I buy more diesel fuel than anything else from this supplier. So, now that gets interesting because the content of a watch list doesn't even have to be a number. Well, wait a minute. Why does it have to be a badge? Why can't I grab data and visualize it in some other forms like a pie chart or a bar chart?
Or what if because we know orchestration can do this. We talked about JD Edwards integrating out to the digital economy, right? And cloud services and thirdparty systems. What if the content could come from external sources?
And here's an example. What if that external source was the National Weather Service? We know orchestrations can go out and do external REST API calls. Orchestrations can provide a input to a thirdparty system, right? Or sometimes we call that a prompt and that thirdparty cloud service does something, returns us something back. In this case, the National Weather Service gives us a temperature. Hey, it's 66 degrees in Denver. And maybe a little uh weather forecast in the form of a text box. So, what if we could chart that in in the text box?
Well, now we've got some interesting data that's coming from JD Edwards and some other systems and we've got some interesting visualizations of that data.
What's the missing piece? Ah, we need to publish those things somewhere in the JD Edwards user interface where our users can get benefit from them. So, what if I could take these visualizations and publish them on Enterprise One pages? And congratulations, tada, we now have Enterprise 1 widgets.
And that's why, you know, it's kind of like we took all the benefit of watch lists, took a good thing, and in my opinion, made it better. And it's pretty simple. You create an orchestration that generates some output, the output that you want. Then in orchestrator studio you can associate that orchestration output with pick your favorite most appropriate widget visualization. Is this a pie chart? Is it a bar bar chart? Is it a text box? Is it a badge? And then you take that widget and you uh it's available to you on an enterprise one page.
And by the way, the badge type widgets, the the the oval boxes look and behave just like watch lists do, right? So they they show up in the little watch list drop-down flag and you can put them in a watch list pane and you can put those on the tiles and connectors between the tiles. So the the badge watch list are really sort of a twin cousin to watch lists and that gets pretty exciting too w with what we can do.
So uh I'm excited about widgets and they're available as of last uh two weeks ago when nine tools 9293 was available.
Live Demo: UX1 Pages and Process Models
All right. That's right. So, let us proceed with taking a look at some of this stuff because that's way more interesting than you're actually gonna do a live demo.
I'm going to attempt a cardinal rule never do a live demo, but we're going to try. So, is widgets a new baby? AJ is a new It's a cousin. Born in April of 2025. Yeah. Born on. Yeah. Yeah. Let's see what's going to happen in the um little sister. Yeah. In the next. Right.
So hopefully I mean this is this is familiar to you. Hopefully this is a typical JD Edwards UX1 role-based page for production manager. And by now hopefully um you're hearing the mantra of UX1 because it's been going on for what eight eight nine years now. It's alert, analyze, act.
And the alerts are in the lefth hand side in the form of that watch list pane. I've got four alerts going on here. They all for my demo data in this demo system, which by the way, I'm running. This is the JD Edwards trial edition running on OCI. So, I love this environment. It's a it's quick, it's solid, it's convenient, it's a fantastic sandbox for me to play around with, and it's available for you, too. So, everything you're seeing here really is JD Edwards running on OCI in that trial edition.
Um, but here I've got four alerts, u, and they're all red. These are WA typical red watch lists, right? So that's fine. The analyze piece are my or you know UX1 analytics charts in the middle those four charts. Uh and on the right is my act and those represent interesting things that I might want to do to launch out if if I want to do something about this situation, right? I might want to go to shop floor workbench or costing inquiry.
And so the alert analyze act is a pretty good paradigm. Um, but what it's missing is, as I said before, the process context, right? These alerts are great, but where are they in the process?
And so the first thing that we did in our um enterprise automation quest is we took the capabilities of enterprise one pages and we said well we can build out these tiles and make it look like a process right and here's one that's you know might be my happy path my very simple order to cache process and I don't pretend to oversimplify the order to cache process, right?
But at some point, you need an umbrella view of what the happy path is because one of the features that we put in that we added to Enterprise One Pages is the ability to click from one page into another page. And that's significant. That's important because if my order to cache process is too complex to fit on one page, I might want to break it into two, three, four, six, 10 pages. and and now I can do that, right? And um so that was one capability that we rolled out uh about a year ago and that's significant.
The other capability was to be able to put watch lists on top of these tiles. Well, now I have those same exact watch list that I might have been seeing in my UX1 rolebase pages, right? my alert analyze act but now they're in the context of hey I can see now that I have 24 orders on hold and that's red because it exceeds my threshold or I have zero prioritized orders right maybe that's a good thing right so it's the same old good good watch list that we know and love um but now they're in the context of my business process.
Well there's a few other things here that if you're looking closely You might um say, "Wait a minute, how did I do that?" One of them is the ability to put those watch lists on the connectors because some of these metrics that I'm interested about don't really apply contextually to the node. They apply to the connector in between the node. So now I can do things um interesting that represent the transition or the relationship between two adjacent nodes. And that's um um that could get interesting.
Uh another thing that you might notice is like we talked about how did I get text in my watch list? Well, this isn't a watch list. This is a widget. And this is just some random text. Maybe you want to put it there. And the the way that gets there is there's an orchestration running in the background to generate that text. The same with this number. Wait a minute. I thought watch lists could only be integers that represented a record count of grid rows.
Well, now with widgets, that orchestration can go and get me a total of all my sales orders. If I'm going to print this invoice, right, I'm going to how much revenue is stacked up behind this activity. And this is an orchestration running in the background that says that uh you know $23,43823 is is is what I got going on here.
Um, another thing you might notice, and this was a a customer enhancement request, by the way, the ability to just drop any image as a tile on my enter one Enterprise One page, or the ability to put some some random text, um, some userdefined text on there as well. So, um, or, you know, hey, uh, can I put some text in a badge on top of a on top of a tile?
So you know think of essentially you know with the UX1 rolebased pages you had a pretty good mechanism and enterprise one pages composed pages in general you know I think we gave you you know a Crayola box with eight crayons in it and you can draw some pretty pretty pictures with eight crayons in my opinion this is you know a crayon box of 64 crayons um lot more tools lot more capabilities again you know we're talking about possibly building or reusing UX1 role-based analytics in the context of these process models.
So now I have an order to cache process model with models and metrics superimposed and I can you know again click on these and link out to other enterprise one pages right so if I want to get you know more detail in my back orders um this is uh content that we ship um and so it's it's really easy it's no code zero code to put these together and I think business analysts are going to look like heroes to the business as you start to sit down and do these and you know trusted partners like ERP suites is going to have um a lot to say with workshops and putting these things together. That's my my prediction.
Widgets and Orchestrator Studio Demo
Over to widgets, right? So let's you know drill into widgets. And I said, "Hey, wouldn't it be cool if we could write an orchestration that generates output that then gets visualized in an appropriate way and then gets published on an Enterprise One page."
And so, um, here we go. This is a widget pane that I'm playing around with. As a matter of fact, all the widgets on this pane are built using a Oracle by example tutorials that are now published on learn JDE. So we can go through and get this purchase orders by business unit and we see purchase orders by business unit is is backed by an orchestration and it looks something like this.
Let's um I said uh we weren't going to of course I timed out live demo uh live demo of of Orchestrator Studio and when you take tools 9293 you're going to magically see this new widgets button here and this is where the magic happens and creating widgets. Um it is there.
So, let's see. We got our um let's say do a sales uh sales analysis. Oh, that's just a text generator. That's not the right one. Sales by business unit, right? And so we first define the w the widget and we say what type do we want? It could be a badge, a meter, a chart, a text box. And if I choose chart, a chart type, I can even do a bar chart, line chart, area, pie chart. And so this widget happens to be a pie chart.
Um, then now that I know what I want my widget to look like, this is where I click my data source. And behind this sales by business unit is an orchestration that's running. And I can click to even drill back quite easily into that orchestration. and say all this orchestration does is a data request to go fetch some data.
And for those of you that you know know orchestrations pretty well, I got some inputs here. My order type is a sales order. My line type is S. Order date from and to dates and which company I want to go. And these are all variables. These are all inputs into it.
And when I run this orchestration, we know how to this orchestration is going to return a bunch of data, right? The sum of the extended price for this business unit. And here's the name of the business unit. So the sum of all sales orders for business unit 30. Um the sum is $129,000, $68,9696. and the cost of those. I can also pull the cost out of the tables.
Well, this array of data then becomes the input the input into the widget that is destined to be visualized as a pie chart. And again, we can bring those inputs from the orchestration even back into the widgets. So if I turn these switches on, it allows me to change and specify the order type and the line type. For now, I'm just keeping all these default values.
But when I preview this, the orchestration runs, takes that array of data, and generates a pie chart, right? which then eventually I can publish on my enterprise one page. And I think you guys kind of know how to do that already, right?
Uh if I go into uh manage content composed pages, now I'm in the noode way to publish all of this stuff on my Enterprise one page. And if I want to add a new widget in here, I click the plus sign. And the magic behind here is I have a new tile type of widget and then I pick the widget I want to put on there. It's as simple as that. It's one, two, three. It's and of of course you know me, I'm oversimplifying everything, right? But it's create an orchestration to generate the data that you want, pick the visualization and build a widget and then publish this widget on an enterprise one page.
And we got some more ideas here. I mean, the cool thing about this, you know, as I said, it doesn't even have to be JD Edwards data, right? This is a National Weather Service. This is the live, you know, forecast in Denver right now. It's one degree cooler than it is in New York City.
Um, here's a steel market price. So if you're interested in, you know, getting commodity pricing, this $31323, uh, that is, you know, that comes right from the Bureau of Labor Statistics. Here's the website where I'm invoking their data service and and getting that data live. So imagine all the possibilities of going out to um um to thirdparty data sources.
So this is widgets. This is using Enterprise 1 pages to build process models. But remember there's one more missing piece, right? What if I want to take the JD Edwards data and automatically build a process model?
Right? We talked about order activity rules and here's uh order activity rules. You got two minutes, AJ, just to give you a heads up. Got it. Got it. Yep. because this is the big crescendo, right? Yes, I know. Here's my order activity rules. Um, purchase orders of stock type. And this this sort of implies a process, doesn't it?
Well, wouldn't it be nice if I could take that? Let's do procure to pay. take all that data and have the system automatically generate it for you. So the same order activity rules starting with enter purchase order and I go to my approval process. I can go to print purchase order. I can go here um that's what this is the enterprise process modeler. This is available with the release 25 today.
Um and this is the thing that's not only drawing the process model for you but also exposing metrics about that right all these come back. This is the amount of open $17,574. Um that's the order value. Um and and I can even say well if if I'd rather not the order value but how many what's the quantity of line items open there? Oh that's 180. Oh, but I'd rather go back and see the the value. Oh, it's 17.6K and the actual real value is 1757430 cents.
So, all kinds of metrics that are superimposed here. I've also got analytics out here. So, I can blow out these charts and say, hey, what if I want to drill into spend analytics by business unit? Looks like Eastern Distribution Center spends most. That's my uh demo data, of course. in the southern distribution center is much less. So I've got analytics adorned there.
Enterprise Process Modeler in Action
There you see there's lots of zeros in here. The zeros mean that the order activity rules allow transition from one step to the other. But remember this is coming from my actual data. So what if I go in here and turn off the zeros to get a clearer picture?
Well, now look at this. um without seeing the zeros, I get a much clearer picture of what my business process actually is. And this is where we think not an end user, but a process owner is going to spend a lot of time in here playing with filters, right? Hey, what if uh you know, what if I want to see a different company? What if I want to see a different time span? What if I a different time range?
Um, we've got some filtering in here and we think that business process owners are going to spend time looking for blind spots. This is what's going to tell the story.
Uh, another nice feature is I can take all of this data and create a snapshot of it. Um, uh, we'll call this a snap of April 30th. Right? And now I've got a snapshot of this data. Why is that important? because now I have something to measure against a month from now, six months from now. I can go back and call this snapshot and compare this to my live data or compare different snapshots to each other.
So, we're you can see why we're very excited in the context of enterprise automation what this new enterprise process modeler can do. Uh it's available. Uh we build these off of templates. We have templates available for order to cache and procure to pay, but we have plans for more of those. We'd like to hear from you of of what the important ones are.
And that is the enterprise automation story. With that, I'm going to wrap up and hand it back over to you. The live demo work done. Thank you, AJ.
And then I'll grab the share again and then I'm going to pass it off to Pashant to bring us home. And obviously this is a AI week. So he's going to talk to us and show us JD Edwards and what we're doing with AI.
And so I'm going to call him the godfather of AI right now just because just cuz we called AJ the uncle godfather pashant. Take us away for artificial intelligence. Go for it.
Yeah, you got to go on side. See there? Now it should work. Thanks people. We fixed it now behind the scenes. So pashant, now we'll be able to hear your mic. Go ahead and talk and people can come. Okay. So I'll go over all over again.
Yeah. Thanks. So I'm here to talk about uh artificial intelligence which is something that's there on everybody's mind. We are continuously reading about it, hearing a lot of things about it, seeing a lot of things in action about artificial intelligence. And artificial intelligence has grown leaps and bounds over the past several years with wide range of applicability and a newer use cases that are coming through every day in and day out.
It has also been very game-changing for a lot of industries and there are newer and newer models being released by multiple organizations makeing the frontiers of artificial intelligence much larger. and with these newer models come up with large of lot of newer use cases and applicability for various industries as well. So this has driven a lot of interest in the space of artificial intelligence and there are a few reasons why it is of greater interest for business applications as well.
The first one being better productivity. Artificial intelligence allows you to automate manual and tedious tasks to improve your overall productivity thereby uh making it much more easier and simpler to do tasks that were mundane, repetitive and time consuming earlier and it also prevents any kind of human error.
The second aspect is about improving efficiencies. Artificial intelligence allows you to achieve more with less and you can process a large amount of data quickly and get deep insights much more faster. thereby enabling much more faster decision making as well.
The third aspect is about driving higher levels of innovation. Artificial intelligence is going to play a key role in the future of innovation in organizations and uh it is also possible to harness your data in ways that were never possible before with the power of artificial intelligence. This leads to significant amount of savings in terms of cost and labor. And this saved cost and labor can be invested in your strategic initiatives that are there in the organizations as well.
Last but not the least, AI can provide better customer and user experience. AI can redefine and reimagine the way user experience is with your business applications and it can advance and simplify the overall user journey. Customers are getting used to using various forms of AI in their day-to-day applications and they're expecting the same kind of behavior with their business applications as well. And this is where AI becomes important for business applications as well. And it also helps to provide a seamless and consistent experience to customers on a day-to-day basis.
JD Edwards AI Enablement Strategy
Now let's talk about the JD Edwards uh enablement strategy for AI. So JD Edwards for is following an enablement strategy for AI through which we are empowering our customers to benefit from Oracle's investments in the intelligence capabilities that are available throughout the Oracle technology stack.
Uh whether it is available through the UCI services or the AI that is embedded within the SAS applications or something that is available out of the box in the Oracle database or on Oracle cloud infrastructure. With this enablement strategy, we are trying to provide ways by which customers can leverage and benefit from these AI capabilities that are available throughout the Oracle technology stack.
We do not plan to embed AI into the data product. Thereby we are providing a more enablement strategy that we're following.
So let's uh dive back and look at the JD Edwards digital platform. I think John and AJ touched upon it quite extensively about how digital platform is being the enabler for various types of things and over the past several years digital platform has been a key enabler for digital transformation for our customers and it has evolved over the past several years enterprise automation being the uh recent one which the digital platform uh is enabling and it is at the core and foundational piece of the enterprise automation layer as well and the digital platform continues to the foundational piece for the next wave of innovation which is the AI wave.
Can you move forward? Yeah.
Yeah. That so it is it'll be key and foundational for the next wave which is the AI wave and we will discuss more in the subsequent slides of how the digital platform has now been enabled to support your AI initiatives this way.
So let's look at what Oracle is doing in the space of AI and how Oracle is has been investing in providing AI throughout the entire technology stack. So Oracle has uh provided AI in all the different layers that it provides on Oracle cloud infrastructure right from the infrastructure layer to the data platforms the AI services and the applications.
Uh at the core of everything is the infrastructure. So the AI workloads require very high high performance infrastructure which is provided by Oracle through the flagship supercluster with RDMA networking that provides high performance infrastructure uh that is required for running AI workloads.
Along with that it also provides HPC file systems that provides high volume storage for storing data that is related to your AI workloads.
Moving further up on the data platforms layer, AI is embedded on the various data platform offerings that Oracle has.
The Oracle database 23AI has the vector search capability that allows you to look at unstructured uh data and also be able to query those. The select AI capability that is there as part of the autonomous database helps you to perform natural language querying on your data that is there in your database.
Uh machine learning is also embedded in the Oracle database along with the data science capabilities that it that provides end-to-end managed uh services for managing and building your data science models.
The data labeling service is also something that is important and foundational for labeling all the data that you're using for your AI workloads so that you can consume it later for u uh any of the AI workloads and processing of AI data.
On top of all this, it's the big layer of AI services. See, so Oracle is democratizing AI by providing a host of AI services uh custommade and pre-trained for specific uh use cases.
So starting with generative AI that provides uh access to large language models that enables you to generate and summarize text and also provide retrieval augmented generation through the generative AI agents. The digital assistant provides a more chatbot-like interface through which you can interact with your data.
The OCI speech service enables you to translate your speech into text. And the OCI language service allows you to uh identify the language and uh translate and pick the key value pairs in the language uh text that is provided uh in a particular language.
The OCI vision service allows you to identify and perform image analysis and identify objects in in images and the OCI document understanding service enables you to process documents and identify key value pairs within the documents.
So these are AI services that are available. These have pre-built and pre-trained models. So these can be used as a much more plug-and-play without requiring a lot of expertise in AI or machine learning. And it also can be customized based on specific customer requirements as well.
And on top of all this, it's the entire application layer stack where uh you have these fusion applications which have the embedded uh AI into that and other third party applications that can consume these uh entire infrastructure uh data platforms and AI services.
Orchestrator Integration with OCI AI Services
So as part of our uh enablement strategy for Oracle AI, the very first building block that we delivered was the authentication of OCI services through the orchestrator. Um AJ briefly touched upon this when he introduced the digital platform. So this uh has been available since tools release 9 to82.
Uh and uh what we've done here is there is a new authentication mechanism called the OCI API keybased authentication that was introduced in the orchestrator so that it can authenticate to and invoke a wide range of OCI services.
So Oracle Cloud Infrastructure or OCI offers close to 150 plus different services with various capabilities. And with this new capability in the orchestrator now you can connect and authenticate and connect to all these 150 plus AI ser uh OCI services including the AI services uh through the rest API connection and uh be able to benefit from those.
So uh uh extending to what I said in the previous slide uh now the JD Edwards orchestrator enables you to connect to these OCI AI services as part of this entire authentication mechanism that we've introduced and this is a foundational block for our enablement strategy where you can connect your orchestrator to AI services send your JDwords data to the AI service the AI service processes that data sends the process data back to your JD Edwards and you can store that for further processing in your JD system.
So this mutual communication between orchestrator and AI service is now enabled by this.
Uh moving on, we'll just uh briefly touch about some of the uh services that we have in the AI space.
First is OCI document understanding service. This allows you to process documents, extract key value pairs and text recognition and uh table detection etc.
Some of the use cases where this can be used is in terms of uh processing your invoices, the receipts in the procure to pay process or adding receipts to an expense report.
The next service is the OCI vision service. So this is basically the computer vision capability of uh AI that uh is provided as a managed service uh which has pre-tra pre-built and pre-trained models as well as customizable models.
Uh there are several use cases where this can be used. First is uh product defect identification. You could train the vision models on your products and uh if a picture of your product is provided to the OCI vision model, it can identify whether the product is defective or not.
You can also manage your inventory by analyzing the stock levels that you have using pictures and videos of your warehouse. You can also identify location of items in your warehouse uh using pictures and videos processed through vision models so that you can quickly access and uh uh provide those items.
You can read product blueprints and also create a bomb uh using OCI vision and uh you can also use it for crop monitoring where you can continuously monitor your crop through through video and images and uh it can identify quickly if the crop is getting in infested or if there are any things that you need to take action on.
Last but not the least, you can also use it for waste management and pest control. So any kind of infestation can be identified using OCI vision.
The next service is about uh OCI generative AI. So this is uh uh uh Oracle uh integration with uh the large language models that are available out there in the market. We've partnered with Meta and Cohair to provide their pre-built models within the OCI generative ser AI service as a rest API.
So this is fully hosted in OCI and there is uh no cross region or crosscloud communication. So it's highly secure.
Some of the use cases where generative AI can be used as for uh document summarization and also retrieving information from documents using retrieval augmented generation agents.
AI Use Cases for JD Edwards Customers
So these are some of the use cases that we've seen uh uh that resonate with our customer base.
The first one being the invoice processing and expense reporting where uh you could use the OCI document understanding service uh to process uh invoices and uh expenses because this is a high volume manual and tedious process that uh is most common across organizations and uh AI can help you automate this process and uh benefit from uh saving time and energy.
The next one is about order capture. We have uh one of our customers where they're taking orders uh the sales reps are taking orders by paper on paper and verbally as well and it becomes very hard for them to put that back into the JD Edward system and transfer it into a digital format. So we're working on how you can use a combination of or OCI document understanding and speech to text to convert these orders into a digital format without having to do that manually.
And uh the last but not the least the uh uh AI use case here is uh how do you make static data that is available on your reports to a much more dynamic and interactive model. So uh c the customer is exploring usage of a chatbot to access the JDwords data that is available from order details approvals and product availability etc. using a combination of Oracle digital assistant and the OCI language service by which you can interact and uh get the data through a chatbot like interface.
So I would like to also u uh let all of you know that we have a JD Edwards and Oracle AI page there on learnjd.com. This is available right now and where we posted a lot of artifacts that can help you get started with your Oracle uh AI journey.
So we have uh some of the documents like business brief, technical brief uh and FAQ documents. We also have some examples that we've posted out there uh which you can follow like step by step and be able to play around with the Oracle AI capabilities along with orchestrator.
Back to you John.
Thank you.
Excellent. Thank you, uh, Pashant. Uh, just a few more minutes, so we'll wrap it up. If you have any questions, go ahead and put those in Q&A real quick. We saw one question that AJ already answered. There's another question I see, uh, from Kylie about the JD anniversary. I'll get to that in one second.
Closing Q&A and Wrap-Up
But learn JD. This is where all the uh, if you want to find out any more information about anything that we talked about today, go out to learn JDE. Uh, like I said, it's good to do these virtual conferences, but it's great to see people face to face. So, I can see your reactions in the audience versus right now I'm just talking to a vacuum. I can't even see besides what you guys post in the chat.
So, come see us live in person. Pashant is actually going to be there in person this year flying over from India. AJ and I will be there as well at the Venetian June 9th through the 13th.
I mentioned earlier the JD Ever Connection podcast. That's Paul Hal Cooper on the right there and Chandra Wabshaw in the middle. They started a podcast in January of last year about 50 or so uh 30 minute sessions they do where they talk anything and everything JDE designed for business analyst CNC's great way to get information and so that's what we have for today lots of content we know it's a busy week of a lot of content stay connected here are various channels.
The question that Kylie asked does anybody know that what is the JD Edwards anniversary I gave you the year but who knows the actual date the date the year is 1977. The actual date is St. Patrick's Day, so March 17th.
So, we just passed I'm not good at math, but I think it was just had our 48th birthday. We are looking forward to 2027 when JD Edwards as a company turns 50.
So, one minute to spare. Any other questions? Kylie also said it's an ugly baby, but we will we let her ask the question about JD Elver's anniversary. And by the way, I put that question in chat GPT and it did not know the exact date. So we are still needed in this process. We are still human intelligence is important because we have that chat GPT knew the year but it did not know the actual date. There you have it.
Anything else? No other questions. Yeah, actually William William William posted one, right? He said, "Hey, I got got a lot of watch lists. Um, and and the performance sort of suffers and you think about what's going on in the background."
And with widgets and publishing those, you can imagine it, you know, it only exacerbates that. So, um, yes, you're going to definitely need to, uh, get some architecture help and make sure that your system is scaled and tuned to be able to handle all the digital gold that you're pulling out of it.
So, yeah, don't underestimate the value of of uh, good system architecture, scalability, great reason to go to cloud, right? elastic resources. You need more, get more. Um, and I know personally that some of the best system architects are right around here.
Oh, yep. So, that is a wrap. Thank you all for attending. Enjoy the rest of the sessions. I see that there was a poll about going to release 25 and we see a number of customers are planning to do that in the near future. So, we looking forward to seeing those go live and helping you be successful. Thank you all. Have a great week. Appreciate it.
Topics: