How AI Is Making JD Edwards Smarter: Real Use Cases from ERP Suites

Transcript:
[...0.1s]Why ERP is look like that transactional system? And I think the No. 1 point is the legacy of where ERP was was born, right?It was born [...0.6s] specifically way back in the day, probably before you were born. It [...1.3s] was like I was born by that time though, for sure. [...1.0s] Is your JD Edward system just a data entry machine? What if it could actually help your business think, optimize, forecast, spot pricing gaps, and even prevent downtime?In this episode of not your Grandpa's JD Edwards, I talked with Manuel Nera from ERP Suites about how AI and orchestrator are transforming JDE from a transactional system into a smart, self improving enterprise engine.Stick with us, [...0.5s] you'll walk away knowing how to prep your organization and your data for the next big leap in the RP. [...5.0s]Welcome back to not your Grandpa's JD Edwards. I'm your host, Nate Bushfield. Today's episode dives into a bold question, can JD Edwards be smart? To help some practice? I'm joined by Manuel Neira, [...0.5s] VP of AI and products at ERP Suites. Manuel has decades of experience at the intersection of ERP product development and AI innovation.Literally from JD Edwards to Oracle and now into intelligent automation. We're gonna talk about what it really takes to introduce AI into your JD environment and why it's not as far fetched or futuristic as you might think. Manuel, welcome.First off, how are you doing today? Doing great, Nate. It's a [...0.6s] sunny day here in Denver. Bright blue skies and we're in full swing of spring, although that also brings a lot of variability in Denver with its weather. But right now it's beautiful. But happy to be here. Thanks for, for inviting me, Nate.Yeah, of course. I mean yeah, we finally got a sunny day here in Cincinnati. Everything's flooded. It's very, [...0.5s] it's a weird time. I saw somebody swimming in the street earlier today, that was a really weird experience. He was on top of somebody's car that's mostly submerged, [...0.5s] very, very weird look.Um, but for those who haven't met you yet, can you share a bit about your background and maybe what brought you into the world JD Edwards [...0.6s] absolutely.So, um, [...1.0s] I start off in JD Edwards in 1995 as a software developer straight out of college. Uh, I went to university of Colorado school buffs, [...1.4s] and, you know, prime there, you know, being there is brought us back on the map in the football scene.So [...0.5s] great time to be [...0.5s] see you buff for sure. But I [...0.5s] was recruited at the university of Colorado and joined the [...0.4s] JD Edwards at the time [...0.4s] to help build what was Project Everest.That was the internal project name [...0.5s] for what was then, you know, we're gonna about to release this one world that we did in, kind of, the second half of, of 1990s right towards the end of 1990s, [...0.6s] uh, brought that up, you know, with, with the team developed functionality.I was in the, [...0.5s] in the financials area developing the financial set of apps, and then grew through a variety of different roles that I held [...0.5s] application manager, running product management, product strategy for different areas, [...1.0s] went into the development [...0.5s] of rapid Start, which you call them engineers.And Oracle terminology [...1.0s] was Oracle Business Accelerators, was the term versus rapid start. Then it worked with the partner ecosystem [...0.7s] and got the travel over the world. As, as [...0.8s] you know, Julia Woods grew. We got acquired in 2003 by Peoplesoft, then 2005, 18 months later by Oracle.So yes, but now it's been a long story. Now here I am [...0.5s] Eddie RP Suites still based in Denver [...0.7s] enjoying it quite a, quite a bit. A lot of cool stuff that we're doing with AI and helping our customers transform their business.So it's a little bit of a nutshell big, big football fan. As you can tell, big sports fans [...1.0s] definitely love the sun and snow. Yeah, gotta love Coach Prime. He's doing crazy things out there in the sun. Ah, let's not make any comments about him.I, I can't see him going in the top three, but I mean, there are some realists out there that are with me. There are some like his dad that believe he is, but let's just leave it at that. [...1.3s] But we will see, we will see how that plays out. I get where you're coming from though. That's true.All right. Well, [...0.6s] let's get into the reason why we brought you on here. Let's start with the foundational truth. Most companies are still using JD Edwards primarily for data entry. Why do you think JD Edwards has traditionally been seen as a transaction only system.And, you know, so that's a great question. And I'm actually gonna abstract it a little bit broader.I'm gonna say why ERP is look like that transactional system. And I think the No. 1 point is the legacy of where ERP was, was born, right? It was born [...0.6s] specifically way back in the day, probably before you were born. It [...1.7s] was born by that time though for sure, [...1.5s] but it was born for to be an accounting, you know, system of record. That's what it was, right?So it was, [...0.4s] it was meant [...0.4s] for keying information and then [...0.6s] as companies adopted ERP from their financial perspective, there was other needs, right?Other needs in different parts of the businesses, procurement, other manufacturing, and then it grew into operational areas such as sales, right, and [...0.6s] CRM, etc. So it started expanding.So, but the origin was a transaction based system and people that certainly grew up in that, maybe my generation, even the generation before me, right, looked at it that way. But the reality is that it's evolved. It's evolved through time.So, but the big point is its legacy to they are the investments that customers have made for ancillary systems that surround, er, people.Whether it was a preference for certain capabilities or the capabilities that existed in the ERPB, JD Edwards or other systems, right, did not mean their needs. They bought ancillary systems [...0.4s] and now they have that.So they continue to look at [...0.5s] JD Edwards [...0.8s] as a transactional or other earpiece as a transaction system.But I really, I believe in today's age cause we're here, right? 2,000, 25, 30 years since I started this industry. It's just crazy to say, [...0.5s] but here I am with a few less hairs and a more, more gray of gray hairs than I did when I started JD Edwards.But I think the biggest impediment or the biggest is humans humans perceptions, humans how we look at things we make, we are judgmental people, right? And we make decisions on technology and it's hard to break away from that.But the reality is I think what you're getting at is ERP is much more than a, you know, transactional type of system or just in entering data. Ready. It can serve a lot of different things and certainly AI is gonna transform that even further.Yeah. So [...0.8s] it's interesting to even think about what it was originally thought of because all these different applications, everything that is really been changing in the ERP space, yes, but also in the JD Edwards space.It's insane how much that it really is involving. But [...0.7s] what would you say is holding back organizations from [...0.6s] evolving [...0.6s] JD Edwards Erps into something smarter, something more autonomous? [...2.0s]Great question. I think it's, [...0.7s] it's a couple of things. It's, it's a little bit of a diconobe here, right? One lack of awareness, right lack of awareness of what real AI is.And when I say real AI, I really mean enterprise AI, [...0.7s] the thing that all of us have probably used in some shapewear form, right, [...0.4s] has been using Chap GBT [...0.7s] to produce an image, to produce, you know, a refined [...0.5s] job posting or, you know, you know, refined a document, a set of marketing.We're publishing a note, no, no digs at marketing, right? But there's certain things that we can do there with minimal risk, but when people start thinking about that type of functionality in the context of enterprise or business, then there's a little bit skittiness, right?Saying well, those are the limits of AI, [...0.5s] but the reality is a chatty BT versus a true enterprise AI solution while related, while having similar technologies, very different types of systems, [...0.6s] right?And you did mention AI, obviously, we [...0.5s] obviously were talking about it, but [...0.8s] when would you say a company is going to know when their ERP is ready for AI? [...2.4s]Great, great question. Um, and it's what one thing is the system which we'll talk about here in the moment, [...1.0s] but again, right, we're still in charge, right in terms of humans, right? We, we haven't gotten to Skynet yet, right, in terms of the Terminator movie and, and that kind of scenario.So it's first thing making sure people understand, people have a strategy, people have [...0.8s] policies in place for how AI is gonna be used and what's not going to be permitted as well. Um, it is very important, right having backing by the leadership team.Once you have that, then it's about, [...0.6s] hey, have you taken care of that maintenance that is necessary, that upkeep [...0.4s] of making sure that you're ready to adopt AI.It's no different that if you're a company and you're running Quickbooks and you grow up and now you need a full strength the RP like JD Edwards or [...0.5s] the others that Oracle has.Um, it's about making sure you have your processes defined, make sure that things are in top shape, and, and you're gonna get the most out of the solution.AI is no different, [...0.5s] right? If you don't make sure that you've done your prep work for data, make sure data's in good shape, make sure compliances outline properly as well as [...0.4s] security, it's defined, who can use it, how can they use it, what features are they gonna have?You know, when, when you have those things, [...0.6s] then, then you're ready to adopt now. It doesn't need to be big bag either, Nate, you can do this these kinds of things in spot areas.And that's actually what we're seeing from customers. They want quick wins. They want to be able to prove that enterprise AI is real, not just from a solution perspective, but actually [...0.6s] real value and being driven into the business.Yeah, [...0.6s] I will say I'm more of an Eye Robot fan. Terminator was good too, though I mean, I'm not gonna discount the part of Schwarzenegger. I mean, that guy, he knows how to act and now he's a, [...0.5s] I was a senator somewhere. I don't know how that works, but good for you. How did that happen?I don't know, I don't know, but [...1.0s] I took a look at your presentation that you recently had at SKUG [...0.6s] and I loved the 4 phase AI Journey roadmap, starting with alignment Day and ending with real solutions and production. Can you maybe walk us through those phases? What happens at each step and mostly why [...0.6s] each step is really needed? Got you, no great point.And overall, what this is what I want to, you know, when I speak to people about the journey that we put together, it's, it's analogous to an implementation methodology of an ERP, right?It's really given a map to see how do you make sure [...0.8s] that ultimately you have a process, how you're gonna implement that, you're gonna cover your bases, and that at the end, when you're finished implementing, you not only get a working solution, but, like I said earlier, you get the return on investment.Because let's be real, right, [...0.6s] none none none, none none of our customers [...0.5s] or companies out there using JD Edwards will adopt new technology unless there's something in it for the business, right. And and, and it's gonna drive efficiencies in value.So that's what we decided to carve out, knowing that each customer has [...0.6s] different, you know, kind of, political situation internally in the company.What are those, the perception of AI, how their system is set up, what they're lurking to accomplish, right, and, and what are the areas that need to be looked at.So that four phase approach, number one is alignment date, really getting together with [...0.4s] leadership at the customer, and really understand where they at right. Have they established a strategy?Increasingly, we see that customers are right establishing a strategy around AI and now it's like, well, what, next, [...0.6s] you know, what, where do we start in terms of use cases?But that first phase is key to really understand where the customers at. And then we have self assessment.We have customers of self assess, we assess to and say, okay, this is where we think you are on the, on the spectrum [...0.5s] of AI adoption, where you're ready or is there some other things that need to get done? And people will, you know, companies will be all over the map.That's first phase.Second phase is really getting deep dive in doing the assessment right? Then, like I was talking earlier, [...0.6s] what's the state of, of our data, right?You know, and the key thing sometimes that's talked about in gt Edwards is the master data, what's the state of the master data and is it clean? Is it is, it is a representative of the entire business sometimes.Yes, I know there are some customers very diligent and, and making sure their data is clean and even so have implemented solutions to ensure that data is clean and consistent. Uh, so kudos, right, not ever is at the same state.Um, there's also security, like I talked about earlier, compliance, right, is another, [...0.5s] another component as well.And last but not least is assessing [...0.5s] the use cases. What are the ones that make sense? Because the use case that might fit your business, Nate, may be a fantastic one for your business, but maybe it's not, it's not the primary one, right?There's not enough friction in that error for me that I need to look somewhere else.So [...0.5s] that's the assessment road nap [...0.4s] and a phase in a nutshell. Next is, okay, once we've assessed at each one of those steps that I highlighted briefly, you will get as a customer, uh, an assessment document highlights what we found, where their opportunities, where things are strong.And it'll carve out, okay, some some work, some homework that needs to be done or housekeeping that needs to be done [...0.4s] with your JD Edward system.Phase 2, then is the cleanse and optimize, right, had, you know, address that, [...0.4s] and we certainly both having intimate [...0.6s] insight to the customer's environment ideas.We can help with that, but customers have the option if they want to do that themselves. They can, right, when we actually, you know, that's why we give the the, the documents that we provide [...0.5s] with our findings and what we recommend, and they can implement their own.And then the last phase, last but not least, is actually developing and implementing the AI solution [...0.6s] on, on a solid foundation.I I, I like to use the house analogy, right, is like, if you have a 2. 2, you know, story house, and you wanna build [...0.4s] a third story, you're gonna [...0.4s] make sure that your foundation is strong, right?Oh, yeah, that's no different with, with implementing enterprise AI, which it eat or you're gonna wanna make sure everything's your processing up to snuff in the other areas, just not before you implement. So that's, that's the four phased approach.One thing I do wanna say though, and it's important is [...0.8s] sometimes customers look at, okay, no, this is good, this is very complete, but I need to show something quick, right, I need to show something quick to the business that proves enterprise AI is real in terms of value and, and benefits [...0.8s] and the tech, improve the technology, right, get over, kind of, some of the [...0.7s] perceptions. We can do that as well.We don't, you know, it doesn't have to be Big Bang. It could be in a spot area where we do a compressed version of this.Yeah, of course, and obviously in those early phases, you were talking about data master data. Um. So in terms of the actual security and maybe even the data governor governance, [...1.0s] how important is that in terms of those early phases?How important is it that, that is under control, [...1.0s] extremely [...0.8s] cannot, cannot imagine, right bit?While [...0.8s] as we know, AI can do some very interesting things right there in some of which we only get a little kind of sniff of when we're doing chat GPT [...0.6s] what we can do in terms of being able to, [...0.5s] you know have an agent [...0.6s] assess [...0.6s] situations, learn from it, and be able to predict scenarios or recommend courses of action. [...1.3s]Excuse me, that may that may be a great enabler for you, [...0.6s] but if I have access to that and I'm not supposed to have access to that function, it could be detrimental, right? It could be detrimental in terms of, you know of compliance or socks [...0.4s] type of situations.So, so it's important that, you know you be able to set up [...0.8s] the AI solutions for the people that need certain access to be able to do their job, right? To whatever broad spectrum it is, they can [...0.5s] without getting into other areas that maybe you're sensitive.And that's why that's one of the big differences between just standard AI and enterprise AI. The solutions like Oracle, they have constructs, and they actually just announced a new platform [...1.0s] for building AI agents [...0.6s] about just about a couple weeks ago that they just GA.That's meant to simplify things and be able to facilitate [...0.4s] the creation of security and, and and be able to set up compliance [...0.8s] properly. I mean it's easier than ever to create your own, or at least assess your own security.I mean that's, that's what's great about JD Edwards is still rolling out more creative and cutting edge applications that are there to help their users. Um, but how important would you say change management is when introducing AI to long standing processes and even teams? [...1.2s] It's massive, and it's [...1.3s] increasingly in the AI space.It's not just the technology, and don't get me wrong, Nate [...0.6s] and everyone listening, [...0.8s] the technologies that we look at right using here are very different than what we're used to in the ERP world, right.Being able to [...0.4s] create transactions and in a sales order and processing it, and even having some intelligence and some of those apps to be able to predict, right, um, [...0.8s] pricing with our with our advanced pricing engines and what not.But now you take it to the next level with certain type of capabilities and [...1.3s] the technology while for folks like you and I that are very in very interested in, and we get wowed by what technology can do, that can also be something that, that drives a little bit of fear and uncertainty to an individual, right? What is this really gonna do? How is it gonna do?What are the side effects, right? And the, the other part of it is, [...0.6s] well, I'm going to be replaced right by that AI agent, [...0.5s] and like I said earlier, right, we're not at that sky net [...0.5s] point, yeah, not yet not yet but what is true is there are companies today and that industry is [...0.4s] growing, right?Companies that are adopting AI solutions. And it's in the billions of dollars that people, you know probably in hundreds of billions of dollars now that are investing in AI solutions [...0.4s] and what, you know by real humanity being replaced by machines [...0.6s] far fetched right at this point in time and maybe for, for [...0.5s] quite a bit of time.But what's real today is if you're using AI, Nate, [...0.7s] to be more effective more efficient and I'm not and I'm competing with you. I am absolutely at a disadvantage there, right?In terms of being a adopt. So that is certainly [...0.5s] one thing, right? Change management back to your point, right? I tie this together. Change management from a technology perspective is certainly a component because we're gonna have to learn how to adopt, [...0.7s] but the change management, I think that's the biggest part.That's why methodologies like the one I highlighted earlier are important because we will also deal there with throughout the four cycles is the change management side of the house by people not only getting over [...0.5s] and understanding that maybe the cheese will be moved a bit in terms of jobs low, right? But it'll be enhanced.So some of, you know you and I and I think a lot of people listening to this podcast probably have more work to do than they can get in any day.It's just a reality of today's, today's world, right? So how about a world where you can get your stuff faster and you can get to those actual things that are differentiators. That's what it's really about. It may change some things.So we have to learn how to work with AI to be able to get the maximum out of it and it puts us in better situations to be successful.Exactly. I mean I will say I have, obviously, everyone is probably heard of if you put two AI robots in the same room, they come up with some language, just so they can talk to each other without humans understanding.But this is worlds away from that. There's no reason to be alarmed. Um, well, but let's talk about actual AI solutions that you build ahead of your peace suites, cause some of them some of them are wild, but in, in a good way, in a safe way.So can you break down the automated quote to order solution that uses document understanding in orchestrations?Yes, absolutely. So [...0.4s] one of the things that, you know that we've been leveraging our AI is [...0.8s] document understanding type of functionality, right, that can read documents right can read documents not only printed documents, right, something that's typed out or electronically and then sent, but it can actually read handwriting.No, and that technology is evolving, right? So I don't wanna I don't wanna it's still growing and maturing, but it, you know we've been able to get some use cases successfully reading handwriting, maybe not mine, [...0.7s] because [...0.8s] I'm not a doctor, but I have doctor like, handwriting.Let's put it that way. [...1.3s] But, but [...0.6s] in the, that quote to order, the, that the process were trying to help with a set of customers, is being able to streamline the process of getting a request for a quote, be an input and getting into JD Edwards quickly and shortening the time. No 1, to be able to create [...0.5s] the quote. 2, being able to optimize the quote.Third, being able to have it approved [...0.6s] by, by a manager. [...1.2s] And that process today was taken longer than expected, and needs us to say it was affecting [...0.8s] conversion rates, right, actually getting a sales order [...0.4s] completed and said, yes, I wanna purchase that.So with, with that particulars, that's it, you know that being able to read the document on automatically putting it into JD Edwards to being able to use anomaly detection, which can use be used for anomalies but can also be used for conformities.You know they'll find patterns that are that are it's successful, right, in terms of being able to get a sales order [...1.0s] approved. So we use that and we cut out a significant amount of time and be able to drive significantly, you know increase in their conversion rates [...0.5s] for their orders.Right? And [...0.5s] obviously we're talking more about value added here or the value lost.And I feel like a lot of people here are either listening to this to understand more about AI or to find out a way that they can add more value utilizing this new, well, new, I say, but AI has been around for years, but, uh, new cutting edge trying to [...0.5s] change their business.So, [...0.7s] so touch on a little bit more about that presentation, but also more of what [...0.7s] AI is doing in ERP suites. The digital sales assistant, how does that fit into JD Edwards users work flows and what kind of value does it add? [...1.1s]Great question. So in actually, that could be an extension of the quote to order process or it could be [...0.4s] a standalone [...0.7s] part of it.And I'm glad you brought this one up because it is a little bit different modality, right? The modality that I talked about with the quote to order is what we're calling embedded AI, right?That, that functionality of document understanding is embedded in the sales order application in JD Edwards. So the user still works in whatever, you know the CSR is working in their home, right? Just sales order sales order entry [...0.4s] and they get everything working there, so they continue.And that [...0.6s] example, you're enhancing that person's experience, making them more productive while still staying in the same domain [...0.7s] now.But in talking to, [...0.4s] you know to other use cases, other customers, the sales assistant was needed to be able to [...0.6s] help CSRS, maybe green CSRS [...0.6s] that are new to the to the business, but need to be productive just as much, right?And, and in cases where [...0.5s] there are cross sell up sell top of opportunities that are more sophisticated, then what even Jr, Jr has some capabilities in that space, but able to do something more sophisticated to be able to look at and be able to look through thousands, if not hundreds of thousands or more records and be able to draw [...0.7s] no suggestions.Plus being able to advise the sales, sales customer, sales representative and say when customers buy this, they also have dependencies, and being able to buy these other things.They're not necessarily part of a kit, it's independent, but it's something that we found that they typically come back for.And being able to provide that insight to the CSR to make him or her a little bit more productive, helpful, right, from the customer service perspective is what it really does, as well as being able to streamline the approval process to be able to do it with a digital assistant.And the beauty of it, you might say, well, now they're outside of the digital system, what happens to JD Edwards? Right?So as the CSR is interacting with the digital assistant as well as the person, you know manager that may be approving, right, the special kind of [...0.6s] look, it's, it all gets updated once it's approved, it's updated real time into data. So again, there's no additional typing, there's no transfer data, right? It all does it seamlessly within the solution that we built.So again, it, it you know that's, it streams, lines it in a couple cases as has enhanced the significantly. Oh yeah, I mean the AI sales, well, the sales [...0.4s] assistant, [...0.4s] I feel like it's [...0.7s] right now, it's at the ground, like we're only making it better, only building it up.There's so much opportunity that can really come with it. But in terms of [...1.2s] the actual identifying of it, how is AI not only identifying issues, but it's also actively fixing and even flagging them in JD Edwards? [...1.3s]Oh yes, and again that, that one's [...0.5s] an interesting one in terms of being able to find [...1.4s] not only in this, this is not only I'll talk about it from the business side of the house, but it's also applicable in the technical side technical world of, of JD Edwards, right?Being able to look at, right, and like system stability and being able to predict things so narrow before as well.As you said, one thing is to predict [...1.0s] other things to advise. A completely different thing is to self heal, [...0.7s] right? So, uh, so that's those doing that. So I think we know we can go on and I know we have [...1.0s] a bit of time here, but I'll speak about the business side of it is being able to look at those transactions and maybe historically [...0.7s] those transactions.This is where it gets interesting in terms of being able to find patterns is saying there is a pattern, but recognizing those patterns actually inconsistent [...0.6s] with more recent patterns.What happened if they go out and find out, oh, we changed some, some business processes [...0.7s] because of an you know because business the business environment changed, competitor that something that disrupted the business, and now we're gonna further disrupt it in some sense of court away, be able to adapt. That [...0.6s] is what's important, right?That old systems in terms of being able to look at just plain analytics, let's say here's a pattern here's a pattern here's a pattern right? Versus AI where I can find the patterns and they can say, well, this one's the prevalent one, right?And where it gets even more interesting as you can look at external factors, right, being able to bring in outside data without exposing your internal precious data. It should be private, right, for your company.You can augment it with outside data and say, hey, this is what's going on in the general industry and it aligns to what you're thinking. This is a good thing, right? And maybe, maybe even be able to suggest some things above and beyond what you've considered.That's where it gets interesting, being able to differentiate some patterns that maybe are dated versus the ones that are new and going to be more effective in today's world. [...1.9s] Again, I mean there's so much opportunity when it comes to AI.There's so much that we're doing as Earp suites. There's so much people out there in the world that are really on the cutting edge of AI and really making the changes. But something that [...0.5s] actually Drew, Rob was on the pod not too long ago.We talked, yeah, and we talked a little bit about AI, talked about how it works with the job structure, how people should be trained for it to better utilize and get more out of their AI solution.So, I mean with [...0.6s] applications like this, with products like this, it really shows how the [...0.5s] combination of both a human mind and AI really can elevate your business to a [...0.6s] higher level and what it can do [...0.4s] just on its own.But if your business is if your business is looking at how to take JD Edwards beyond data entry, now is the time to start the conversation. You can head to ERP.Suites. Com and learn more about ERP Suites AI assessment process or connect with Manuel's team to begin your own AI journey, whether it's one use case or long term roadmap.But that's a wrap [...0.6s] for today's not your grandpa's JD Edwards. Huge shout out to Manuel for helping us understand how AI is not only possible inside JD, [...0.4s] but it's already happening.If you found value in today's episode, hit that subscribe button, leave a review, share it with your team or someone who still thinks AI is science fiction. Until then, we'll catch you next time. Keep asking questions, keep modernizing, and keep making JD Edwards just damn efficient. Thank you. [...4.6s]