AI 101 for Enterprise: How to Start Using AI with JD Edwards ERP
July 30th, 2025
11 min read
In this informative session, Drew Rob, AI Advisor at ERP Suites, walks through the fundamentals of AI and its growing impact on enterprise operations, particularly for JD Edwards users. The session covers everything from what AI is, key technologies, and real-world use cases to how organizations can get started with quick-win engagements. With a focus on education, seamless integration, and practical applications using orchestration, digital assistants, and embedded AI, this session empowers businesses to confidently take their first steps into AI transformation.
Table of Contents
Transcript
Introduction
I'm glad you could join us today. Just wanted to introduce myself. I am Drew Rob, your AI advisor at ERP Suites. I've with, I've been with ERP Suites for five years now, first starting the product space, then moved into more of a data analytics, data consulting role. And then from there ventured into this wonderful world of AI and been in that role for about two years now. And my primary focus is really education, you know, going to user groups and conferences and presenting on AI. Because the biggest thing with, with AI is not just building out products and services, but really being AI advisors and helping you along the way with whatever your journey may be. You may be beginning, beginning your, your steps into AI may just start thinking about tools and technologies used, maybe some different use cases you're thinking about that you may see this week as well and just other various technologies that that we will explore today. But wherever you're starting, our team and others can get you there. And, and really it's, it's all about just getting started and really starting to implement it into your business. So with that, we're gonna, we're gonna move into some more foundational things with AI.
Agenda
So quick agenda today. First, we'll start off from Ground Zero. What is AI kind of what is important? Why is it important? What's the value of it? And then we'll move into really the AI for the enterprise, right? And, and with, with that even talk about JD Edwards using orchestration and really the power with orchestration combined with AI and get into that. We'll also go into various tools and services at a very high level. Now, there are a lot of sessions this week that we'll talk more about these specific tools and services and to go a lot more detail as far as demos, I might highlight a few, but definitely be on the lookout for that. And if you have any questions about other sessions and if you want to learn more about a certain tool or service or or certain type of AI I guarantee you'll be on the agenda at some point this week. And then lastly, the biggest thing is how to get started. How can you get started on your AI journey? And as far as ERP Suites is concerned, we want to be advisors. We have a lot of tools and technologies out there to get you started, but it's really all about starting the discovery process, right? Start thinking about those various use cases and in ways you can implement these tools in your business. So just really tailoring that and understanding AI for the enterprise is probably the most important thing here. So with that, we'll move forward.
AI Foundations
So what is AI? Just a real quick high level, what is AI, right? It it's really just technology that enables computers and machines to stimulate humans intelligence and problem solving. Now, again, that's just a baseline definition, right? But it covers a lot of different tools and frameworks, different applications at ERP Suites. We'd like to integrate with JD Edwards. So right now we're using OCI, but we have also explored in the cloud of AWS and Azure as well. It's really just about finding your best fit. What really fits you guys as far as tools and technologies. As far as I can tell you, and we'll get into it later, some tools are further along than others and how we've kind of utilized different things. We'll go into that further, but it's really just understanding that even though AI has been around for a very long time, it's really in the past, just two or three years really starting to make headway into our lives, especially in the enterprise. And again, it's become more readily available than ever, right? So we talked about ChatGPT, other large language models out there, digital assistants, chat bots, and more than anything it's, it's really just understanding that it's becoming more easier to use. You really don't need a data scientist in your organization to start using it. Yes, you need some underlying baseline education, but these tools are becoming easier to use and becoming more streamlined and easier to understand. So I think that's the most important thing to understand about today that all, although AI covers a lot of broad areas, we're going to help you understand how to really start implementing AI into the enterprise today.
Key AI Technologies & Terminology
So with that, we'll go a little bit further. I just want to highlight just a few of these, right? And we'll really break these down further and how they apply to enterprise later on. But you probably heard a machine learning, deep learning, really predictive analytics and really talking about how to build off of just not just automating certain processes, but how to start making predictions inside of your data, right? And starting to do forecasting, budgeting things along that nature is definitely something we're we're exploring in the enterprise as well as digital assistance and natural language process processing. It's becoming easier, for example, not just writing queries to understand data, but actually using digital assistance to understand your data and using natural language human interpretations to actually perform tasks and ask questions about your data. So all of that is becoming more relevant and easily accessible computer vision. So this one's interesting and we'll go into this a little bit further with some examples, but it's really being able to read images, documents, text, start to understand that side of thing to actually gain data from not just a structured data source, right? It's really that unstructured data. How can we make it readily available and actually uploaded and imported into our certain data services so we can glean insights off of it? The bottom 2 I just wanted to highlight again 'cause they're they're thrown out around a lot machines performing task photography. So robotics processing is another very big one, as well as augmented virtual reality. Again, just generating environments, blend reality. You've probably seen those with VR headsets, but the biggest thing with those two is they're just, you know, going through exploring them with that. And my team, it's not as prevalent in the enterprise today, especially in the work that you guys probably do day in and day out. But just wanted to highlight it as a potential type if AI as well. This one, I'm going to have a visual behind it, but these are just some other terminologies. So genitive AI. So it's really the the way to produce types of different content. And we kind of mentioned it with vision via text, imagery, audio, synthetic data, feeding that into a large language model and actually be able to understand the sequence of sequence tax and, and, and grab data from that.
Retrieval-Augmented Generation Deep Dive
So retrieval augmented generation's a big word that we use. And it's really just understanding via natural processing, grabbing documents from, you know, grabbing data from documents, whether it be manufacturing documents. We, we've run a few RAG models and large language models with financial statements is another big one, you know, and just, and just trying different things, but really harnessing that generative AI solution. One of the biggest things about RAG is, is really just the improved data quality and relevance of generated responses responses. So if you think about it this way, a lot of times in in day-to-day lives, what people are really using is, is stuff just as ChatGPT, three or four other open source large language models. What RAG really does for you is it actually connects to your enterprise data sources, whether that be JD Edwards, you know an Oracle database, other third party sources that might be in ACRM or a, you know other sources like that. It actually connects the large language model to those actually train and get accurate data from those. And then later on we'll kind of talk how how we're using it currently in our business. But first, let's really just talk about the query. So when we talk about the query on that left hand side, that's coming from a customer. Now this query as you can, as you can think about it, it's not actually sequel, it's more of, you know, a natural language response, like what is my inventory count for item number 220 or something along those lines. Just think of that as like high level, that's what it would be. From there, we actually get that response and we look at various knowledge based embedding. So I mentioned JDL was Oracle databases, other uploaded files, financial documents even. And it will look at those various different things, refine the query based off prior input and that being other input from other users, right? It can learn off historical data, run that natural language processing with connected to the LLM and really the LLM is where it will learn from that historical data, connect to that LLM and then send the response back to the user. That's very readily understanding understandable and again, natural language, right easy to understand. And again, with with sending and training these models, these LOL models, it will become smarter as more data becomes readily available. But again, going back to that point before it's very secure when we start to connect these large language models rather than open source models that I talked about earlier.
AI in the Enterprise
So with that, I want to jump forward here to AI for the enterprise. And now this is what I got. I want you guys to really focus on how we're starting to integrate this in the enterprise. So it's really what we really need to focus on is those business challenges you start at those pain points in your business. It's really the discovery. You know, one thing that we've really at ERP Suites that we're noticing is improving customer service, whether that's getting a digital assistant to a territory sales manager or a customer service rep is able to run through training material via a Gen AI model. It's really just about enhancing those business challenges that you're currently facing. So the biggest part is understanding and discovering where those problems arise in your business. Secondly, it relies heavily on your data. And I have it and it should, it's pretty obvious, but it's really all about the data is clean and secure.
You don't want to just implement an AI model or an AI solution without really going back and looking at your master data set up or any other data that you want to connect that model to, right? Whether it's your data inside of JD Edwards, as I said before, or the third party data source, an Oracle database, something you're pulling just as economic data online, it can all be pooled in. But if it's not clean and secure, then the enterprise solution will fail to produce good results, complete results. Next we'll move on to repetitive tasks and workflows. That's the biggest thing. It's really starting with a starting AI with those repetitive tasks, starting from Ground Zero. You'll probably hear it a lot this week, the crawl, walk, run method. Start with automating processes, then move more towards a predictive, analytic, predictive predictability, and then eventually autonomous ERP. But it's all about starting with one simple process and automating that solution. And then again, seamless integration. We'll, we'll talk about a little bit further of how we're using OCI tools with orchestration to make it seamless, easy to understand, readily available, all those sorts of things. That's the biggest thing is you can use a variety of different AI tools, but if the integration in the connection between them all is not seamless, then one, it could cost a lot of money and two, it could cause some latency when you're talking from one mile to another. So those are just some examples there, but really just highlighting these four points is kind of where you want to target enterprise AI when you're first starting out.
AI Use Cases
Enterprise AI use cases I kind of just listed here and I want to talk through a few. So with OCI vision. So when I say vision, again, this is more just reading documents. So when I talk about this, it's about reading expense reports, it's about reading sales orders, it's about reading trying to think of some other off the top of my head, but it's about reading those documents and gaining insights from those. One other area that vision goes with is anomaly detection. So with that it, you can read images of products, right? You send a product through. One example why you like to do is just say a part of a pulley has sent through. What it will do is when you can train it enough, it will actually understand if a pulley is complete or defective and then it will save that data as well. And you can continuously train on various images with these models and it will understand those images better and then learn from them. The second was I kind of mentioned before was Rag, right? That's the ability to really, you know, whether you're connecting to a digital assistant, which we'll talk about later, or just reading various documents, owner's manuals, financial documents, you know, economic resource documents out there that you want to pull in. RAGS is very important.
Leveraging Orchestration with AI
And this is the biggest thing and I’d mentioned in the past, and it's really all about the orchestrations, right? And leveraging orchestrations with AI because especially with that digital assistant examples, orchestrations have the ability to connect the backend systems, really do all the heavy lifting, getting data from JDE, updating JDE, right, making additions to JD Edwards or other sources, right. And again, we, we, we go to orchestration because it's, so, you know, it's easy to understand seamless, but really doing the heavy lifting is the AI and ML tools, right? The heavy lifting of analyzing the data, understanding the data, the historical data that you might put into these tools and really understanding what to do with it. And in the end, that really creates intelligent ERP, right?
Notable AI Tools and Platforms
So with that, I just want to pull up just a few notable AI tools. As you can see, I've highlighted a few. So just open AI is ChatGPT, ChatGPT 4. We talked about earlier about the importance of using RAG with large language models to have enterprise data connected to that. So what I kind of showed here with that second highlight is our ability to connect our digital assistant to a Cohere large language model. And with that, it's a more enterprise model. It's actually an OCI 1 backed by OCI Oracle. So it's more in the enterprise, It's not as open source, it's more secure as what I'm getting at there. Entropic Clod is actually a machine learning, a large language model. We played around with it and it actually creates dashboards for you.
Getting Started with AI
So what we've heard a lot from customers is a lot of times they want to implement AI, but they don't really know how. They're not getting enough buy-in from their executives to do an 8 month, 9 long process of actually implementing an AI solution, which it could take that long to flesh out some of these higher-end solutions. So what's really important is what we kind of coined as our quick win AI engagement. Now what this takes is more of a condensed version of our AI journey, which I'll talk about later. It's really just really defining that use case development and I have steps for building an AI use cases another session tomorrow, but it's really talking about discovery phase discovering. What are your biggest pain points? How can you start to manipulate them in your business? How can AI, what is AI, what is not an AI solution?
AI Readiness & Next Steps
So with that, I wanted to give you a few things to go away with.
At ERP Suites, we have an AI scorecard. What this is an online self-assessment that you would go to and you would, you would assess yourselves on different things, such as where are you at with certain use cases? Where's your security at? What's your data at? Are people in your business very educated on AI? Do they have a strong knowledge on it? Are they starting to accept it, you know, usability or can they adopt it? It's a self-assessment you guys would do. And then we'll have an initial discovery conversation that we talked about earlier, earlier, understand where you are and how we could tailor an AI journey to fit your needs. And then lastly, we'll reassess you. But again, it's all about determining that customer knowledge and AI readiness to make sure we, we can enhance how you operate and how your AI journey goes.
Topics: