Steve  0:00  
What if your wins at work didn't have to stop there? Workhuman's Employee Recognition platform is built on the world's largest rewards marketplace, tying recognition for great work to the things that people love outside it. Crush a project? Snag that baking dish you've been eyeing. Hit a milestone? Celebrate with a weekend getaway. With over 1 million gadgets, gift cards and experiences, something is redeemed for in Workhumans in house store every three seconds. Each reward deepens connections, boosts engagement and helps fuel a culture that knows their work matters and has the receipts to prove it. Head over to workhuman.com to learn more. Workhuman a proud supporter of the HRHappyHour Network. Thanks for joining us.

Steve  0:43  
Welcome to the System of Record show on the HRHappyHour network. I'm Steve Boese. Trish Steed is with me today on System of Record.

Trish  1:00  
I'm in the house. I was excited. I'm like, Okay, why am I on a System of Record? So.

Steve  1:06  
Well you're on the system of record because we were able to get no I want to thank him in advance. Chris Leone, great friend of the show Sunday, HRHappyHour show a number of times, but from Oracle joining us on System of Record on a real busy week for him, lots of announcements going on. AI world events happening all over the place. But Chris, welcome first of all, and then I want to, like, start hitting you up with some really important system of record, system of outcome questions. How are you?

Chris Leone  1:37  
Sure. First great to be here, Steve and Trish Always a pleasure. And yeah, of course, anytime you ask me to come, I'm gonna, I'm gonna make it. So thank you for asking.

Steve  1:46  
Yeah. Well, we really appreciate it, because it is a busy time for sure. And look, everybody's busy, right? But just on the back of a really big announcement from Oracle last week, Oracle Fusion Applications, agentic applications, major announcement, before we talk about the details in the announcement, Chris, you've been writing some really, I'd call them viral LinkedIn posts over the last few months, and the most recent ones are really talking about a real shift in the enterprise technology landscape, from oddly enough, System of Record, which we call this podcast, but to really thinking about those systems differently, into systems of outcomes and systems of execution. So before we get into some of the details and talk about the announcement, maybe, let's start there. What is happening in enterprise tech that's influencing this shift in these big systems?

Chris Leone  2:40  
Yeah, you know, it's been a long time. It's been a long time coming, and I personally and my team have really been very, very deep in the foundational models and the new technology, and we've been delivering on a pretty solid execution path over the last couple years, and, and it's, it's been a really, it's been a really good transition. And, you know, we've, we'll get into more details about about what we've delivered. But, you know, at a high level, the way I kind of describe it is, you know, enterprise systems, enterprise software, has been doing the same thing for the last 30 years, right? I mean, these, these, these systems, have been, you know, huge part of our lives. I've built them for many, many years, but they've, you know, they've been very passive in nature, and that's kind of how I describe it. Look, they've been systems of record. They've been, you know, they've stored our transactions. They've kept them safe. But anytime we needed to move work forward, or there was a cognitive load involved we had to understand the context of a process. Or, you know, had had to get expertise from somewhere outside the system. It was always the human, the person that kind of was the glue for these systems of record and all of the knowledge work that happened outside the system, and that's been the way it we built system for the last 30 years, and it was not because we didn't want to do better, or we didn't want to make the systems more interactive and more knowledgeable. It was we just didn't have the tools to do that. And so that's what I've been kind of writing about, and AI has really changed that perspective and allowed us to do things a lot differently. And AI and agents as what I'll call kind of the currency that we can use now allows us to really design and build systems that really have context, really have expertise built in, really have domain knowledge, and can really help us move work forward in a material way. And, and it really changes the types of systems we can build now. And so that's kind of what I've been writing about. And, and you know, the technology really lends itself to really building a different type of of system, and that's why I say, you know, these systems have to mature, and they have to move from systems of record to systems of engagement, systems of outcomes, which really allow us to execute work.

Trish  5:14  
I'm so glad you said that, because I think that what we're seeing in the market is a lot of just ideation, little incremental changes, and a lot of these companies could potentially make these big changes. What do you think, maybe, before we get into the details, what do you think makes Oracle different? You've spent your biggest chunk of your career there. Like you said, you've been building systems for a long time. To me, it's a fundamental, a fundamentally different way of thinking. Yeah, when you're approaching something like a system of outcomes versus the system of record, what do you think is your sort of special sauce there at Oracle that gives you that push to innovate?

Chris Leone  5:54  
You know, I think,I think we were lucky, or at least the applications team was lucky that we had a lot of technology available to us very, very early. So we were a cloud infrastructure company as well. And very early on, we were given the gift of these large language models, and they were served up to us in a way that allowed us to embed this technology, embed this this intelligence throughout our application suite. Now, we spent a lot of time building our application suite. We really do have an integrated front office, back office and middle office set of applications, and we built them on one data model and and that certainly helped and has worked in our to our advantage. But having a cloud infrastructure component, having large language models available to us early being able to build with the technology. So we started very, very early, taking some early models and building what I'll call generative AI services. And we started doing this two years ago. And you guys were at an off site that we had, and we were briefing the endless and I said, Hey, look, we're going to start building this technology in and we're going to start with these kind of completion prompts, and we're going to build going to build job descriptions that we're going to create goals, and we do all these things with, you know, context from the application and these large language models, and bringing those two things together and delivering, delivering that capability. And then, you know, we progressed into agents and and we said, hey, look, we can go beyond just these kind of single completion prompts in the flow of work. We can create these kind of advisor agents and use new technologies like, you know, RAG and the ability to have a knowledge store and take documents and put them in our vector store and ask questions against them, again, in context with the application and that that allowed us to get to the next set of kind of process improvement. But again, Trish, like you said, it was incremental, right? So the first step may be saved. You know, I used to give, you know, analogies of, like, pennies, dimes and, you know, quarters, and then, you know, these advisors or agents moved to maybe, you know, half dollars and dollars, but it was just kind of incremental improvements and but the big, the big thing that I think our true advantage was it wasn't just my team that was building, kind of the central technology in the AI foundation. It was the large number of application developers that were able to take advantage of the technology, understand what it meant in a supply chain process, in a in a, you know, HR process, or in a financials process, and apply that technology to real business problems that that they were trying to solve. So I think it was a combination of all those things, having technology available, being able to deliver a centralized platform and allowing a large developer team kind of democratize the ability for them to build it into our solutions and not think of it as an add on. So what they were able to do is use it as just another tool that they could add value in the systems that they that they were delivering. They didn't have to worry about, oh, is this going to be a high cost for our customers? We're going to include it and they can just say, Hey, this is how we're going to put this technology in and let our customers get value from it.

Trish  8:59  
Yeah, I love that. And I would add just to, you know, having been a customer years ago, and then all you know, the last 10-12, years, being there and watching all this, because you all have been talking about, AI, forever, right, forever. And it's you hold yourselves accountable, I think, in a way that I don't see other vendors out there in the space, not just in HR space, but just in the tech space, right when you stand up in front of a group and you say, this is what we're going to do, you do it. And the next time we see you, you've done it. So I would say also, I think you're, you're very good with your teams, about just holding yourselves accountable and pushing those boundaries and those limits that you know organizations might have some fear, yeah, right, to do some of these things, and you're just like, nope, we've got it. We're going to include it. You're on it.

Chris Leone  9:44  
Yeah, the whole adoption side, we can talk about the adoption. That's a whole nother story, but, but, but what I can we are having good adoption, but we can talk about, like, some of the struggles that customers are having, and all those things and governance models. But, you know, the one thing that I would say is, like, we've been doing AI for a while, and a lot of vendors are saying Yeah, we've been doing AI. We've had machine learning models. We built our own foundation models. I truly think that the pivot that generative AI and these large language models allowed us to do was in the past, when we built these foundation models on our own, it was a long time. We had to get the data set right. We had to build these models. We had to test them. We had to work with customers. And by the time the beginning of that process started and you got something, you know, truly valuable for customers, it could be a year, or it could be eight months, and it was a very long, drawn out process, and we built a lot of those models, like, like, all of our, you know, the vendors that we compete with, or whatever. But what generative AI did was allow, allowed us to iterate much faster, democratize. We pushed out, we built some central resources, but we pushed out the ability to build these things much faster and in a much more lighter weight way. And, you know, we were rolling out, you know, completion prompts and even advisor agents in weeks, you know, and testing them, and we had big evaluation tests and all those things, but being able to roll out these things very, very quickly and then get customer feedback and then, and then make them better, and it really changed the game for us, and really changed the game for our customers, I believe.

Steve  11:17  
Yeah, Chris, you've mentioned, you've been writing about some of these topics, quite interestingly, on LinkedIn lately. And there's something you wrote about. Was something I was thinking about too. And I did like a 15 or 17 minute Workplace Minute, which is supposed to be like a three minute show, like 17 minutes on this kind of issue couple about a week or so ago. And it was the idea this, and you could tell me what you think about this, and I think about this, and I think I know where you're coming from since you wrote about this somewhat, is some people are saying in enterprise tech, they're seeing a world where these systems of record, whether it's HR, whether it's finance, supply chain, sales, right, all of them manufacturing, they're there. They haven't, as you said, they haven't changed much, right? In 30 years. They're better, they look better, they're easier to use, but fundamentally, they do lots of the same things, right that they did 30 years ago, that they're at risk, potentially right. The value of those systems to customers, and therefore what the systems themselves are worth, quote, unquote, could be really disrupted by some kind of an agentic orchestration layer, or even a whole nother, whether it comes in from a Gemini or anthropic or somebody else builds it to say, hey, just come here. We'll feed in, we'll connect, we'll use. There's some open at there's some protocols now that exist, MCP, there's a couple of others that will help these tools just connect to finance, connect to HR, connect to supply chain. So a user might say, Hey, give me, you know, give me a rolling inventory report and show me the best sales people, whatever it was, right? Just tell the prompt what you want, and the agent, or the agent orchestration engine will just figure out how to go into Oracle and go into whatever manufacturing system you might be using, but something else, and so on and so forth. Right? So on and so forth, right? And spit out the answer, and the user doesn't really care anyway, right? They just want their answer, right? So that's the setting, right? And going to ask a question here, but is that a future, or even a semi future, that we could be heading towards? And because you wrote about that a little bit, you know, about where the value is going to come in enterprise tech. And I love your thoughts on that.

Chris Leone  13:24  
I'm pretty passionate about this. And, you know, I, I like to have a little bit of fun on on LinkedIn.

Chris Leone  13:32  
It's, you know, it's an interesting platform, you know, I like, I do, do like to tease some of our, you know, competitors, you know, I like my work day a little bit. You'll see some me kind of going after Salesforce a little bit, but it's all in fun. Look, I think, you know, I think that the System of Record vendors have a huge advantage. I know how hard it is for us to build agents or agentic systems. And we'll talk about the difference between kind of what we've been doing, from these kind of co-pilots or kind of system type technologies, to truly kind of executing work, which I think is that is truly where we have a big impact on the business, and that's where, that's where I think all of the System Record vendors need to go if they're going to create a moat around their applications. The reality is, the data bricks of the world, and even the sales force that has front office but has to, you know, kind of facilitate back office by taking data out and putting in a big cloud, or, you know, a data lake, all of these and, you know, the service now is of the world that they do have some transactional aspects, you know, they do have a ticketing system. And they do, you know, they do have a service set of products that are very successful, but they don't own the System of Record and and so the what these systems can do is they can take data out. They can kind they can rationalize it. They can build an enterprise graph so they understand, you know, what the different you know how the data relates to one another, and they put a semantic layer on top of it, and you can ask a bunch of questions about it, and you can find out, hey, look who is my next best customer to sell to? So you know, what opportunity should I be pursuing? But the problem is that is gets you, like, 10% of the way to actually doing work for people. And this is what I keep trying to explain to people if you truly want to automate work. And yeah, and I give these examples of what systems of outcomes really mean. If you truly want to automate work, you have to not just say, hey, here are the nine things that I can do, and oh, I'm going to go post a transaction back and figure out how to write back to an API. If you're truly going to automate work, you have to be part of the System of Record, because you're going across an entire end to end process. And I give a sourcing we have a sourcing command center as one of our agentic applications. So so so I think customers are getting value from these assistants that are in the flow of work, that can come from an external system. They're better if they come and they're built in, because you get full context of what's happening and real time ability to update transactions. But if you move to an agentic application, and we're we're delivering now, and where the future will become, is the ability to, you know, to execute end to end work. And so we have an agentic applications called the source of command center. And what this agentic application, can do is you can go, given an outcome that you want to achieve. And this what I tried to explain to you about it, you could go, say, look for the outcome for this process that I want to solve. I want to reduce our supplier spend by 15% percent. I want to reduce lead times by 10% and you give that as an outcome. Then you put together a team of agents that are specialist each one, maybe I have a negotiations agent that knows how to negotiate sourcing events, and maybe I, you know, have an agent that could create great RFQs, and that's what I built it around. It gets content. It understands what you know how to describe that content. It creates these really great RFQs. And I can put this team of agents around this, this problem statement, and they can work and continuously move work forward and create RFQs, create sourcing events, award winners. Now they don't have to do it completely, autonomy, autonomously. And what I what I say is look for now most organizations are going to put human completely in the loop. These, these agentic applications can recommend and say, Hey, these are the next set of actions that you should take. Here's the action that you can take, hit this button and create the RFQ. Hit this button and create the sourcing event. Hit this button and create, you know, award the winners. And there's a bunch of there's 50 different transactions that happen. And it can crank through all of this. Move work forward. Do it 24 by seven, and the human is just saying, Yep, approve, approve, approve. And then at some point you don't need to approve every single one of those. Maybe you feel comfortable with the first 25 set of actions that are kind of lower value with their guardrails, and you just let those happen autonomously, and you're just making the oversight the higher judgment calls. And so what we can do in the System of Record is do all of that work, commit all of those transactions, and do it in a secure way. We still have all of the, you know, functional and data level security, we still understand all the approvals. We have all the auditability. So every thing that happens is fully audible in our transactional system, impossible to do unless you recreate the entire transactional system in a cloud and in a data lake and map all of that stuff. So when I when people tell me that this future of work is a data lake, it's nonsense. They don't understand what it really means to execute work. So now what we can do is we can go to customers, and we can say, you can now have truly elastic labor. You can go and if you have 30 collections people, right? And part of the year you need to scale that to 60 collections people. You don't have to add linear head count and add 30 people. You can make some of that work autonomous, and you can scale that part of the business and then scale it back when you don't need it anymore. So this is where the future of work is going. It has to go in this direction, and you see it with a lot of the different tools that are in the consumer market. I know this is a longer answer than you probably wanted, but if you look at the consumer world, there's the you know, the NEMO clause of the world and the open clause of the world, that these personal assistants, where you give them instructions and they go do work on your behalf, that exact same interaction model is going to happen in the enterprise. You're going to give it work, you're going to monitor the work. It's going to give you feedback, and when it runs into issues, you're going to say, go this way or go that way, and that's how it's going to progress over the next, not five years, over the next, like five to six months. These are the types of solutions now, the adoption curve will be very different. Customers need to feel comfortable with these things. And you know, it's like driving a full self driving car. You know, at the beginning you take your hand off the wheel for going down the block, and then, you know, then it's two blocks, and then you're driving to Napa, just sitting there talking to, you know, your family. That's how this will progress. And so that's how we see the future of work unfolding.

Trish  19:49  
Is that the reception you're getting? Because, I mean, obviously, you just spent time in London, for example, with these announcements. When you're talking to these organizations that are going to be the ones where are using this, are they still at that stage of, I love the analogy in the car, right? So are they still like, kind of like putting their hand on the wheel? Are they like, we trust our relationship with Oracle. You've gotten us this far. We're ready to do a little bit more, because I almost get the sense they might be ahead of some other large organizations that aren't working with Oracle.

Chris Leone  20:21  
I think that I think this is new for a lot of organizations, and the reality is that customers are still adopting, and we have a lot of adoption. We've had 7000 customers adopt our agents and in the flow of work. And so we've had good adoption. I think it will be a long tail for customers to fully move to this kind of, I'm calling it the the autonomous enterprise, you know, I think for customers to kind of move in that direction. But there'll be some early adopters, and the value and the return that they get will be so significant, they'll start to see the huge benefit. And they'll, they'll go through, they'll go through business transformations. The transformations of the future will not be, hey, I'm going to automate this, you know, order to cash process. And this is how that we're going to step through the transactions. It's going to be look, we can look at all the resources that we have within the organization. How much of this do we want to automate? Right? I have collections agents. I have HR, you know, shared service people, how much of their job do I want to automate, and how much of their job do I want to do manually? And how much of it do I want to flex? When I have the ability to flex, that's what the transformations will be going forward. And so that's how system integrators will have to think about it. They're going to say, hey, look, what is all the work that we can automate in the enterprise, and how do we go about doing it? And we're thinking about building the next set of implementations. They're going to go look at that. Here's a list of labor that we can flex and we know how to do it. And some of this can be autonomous, and you can turn it up over time, or you can turn it back over time. And so that's how these deployments will work. And I can speak personally, you know, our business technology coding has been disrupted. I mean, the codex of the world and the cloud codes of the world have really changed how we build our software, you know, and and how we build smaller teams, and how we code things, and how we review things, and all of that has been disrupted. This same disruption is going to happen to all other industries now. It doesn't mean we're gonna have less jobs. Every single technology. Major transformation we have has actually led to more jobs, you know, and so I think it'll just be different type of work that people will be doing. But a lot of the work that can be automated is just going to be automated. I mean, it's just going to be.

Steve  22:40  
Chris, I want to you mentioned system integrators a second ago, and there was one more, at least one more question. I did want to get out there as I was doing some of the research on this and doing some of the pre reading, I thought about that a little bit, right? Oracle, certainly, and some of the other big, big providers work with all these big, great system integrators. You got great partners, right? Go out there work with customers on transformations, on implementations in this, in this world, right in this agentic world, is there room for those groups to still build tools that will, you know, work with these apps in an agentic way? Can they? Can customers build them themselves, like, you know, how can is still, is it open enough where they feel like, hey, I can get in there and create something that maybe Oracle didn't get around to delivering yet, or it's very custom in my organization and but I want to take advantage of these capabilities. I'd love, for your just perspective on that.

Chris Leone  23:38  
Yeah, for sure. I mean, I think there's a there's a couple different perspectives to look at. I mean, I think fundamentally, implementations are going to have to change dramatically. I don't think we're going to have people keying in a bunch of setups and

Trish  23:51  
Can I just clap?

Chris Leone  23:53  
I think, I think we're all, we're all, we're all waiting for those things to happen. So that's happened, and that's happening now. I mean, the reality is, there's so much that can be done from it, from an automation perspective, and and, you know, not having a kind of manual people do do a bunch of this kind of lower level work that that's gonna go away. But, but will customers and and will partners have a have an ability to kind of personalize, configure, even, you know, customized, so to speak, these systems, of course. I mean, the thing is, and the way we've kind of, we've approached everybody will get to execution of work. And this is why I put these shout outs to system of vendor counterparts, because the reality is, every system of record has to go in this direction. AI will become the execution layer for all these systems. It has to. There's no other choice. It has to be. It'll be part of our stack. It will become how work gets done and executed. Period. Just it's going to happen. Has to happen if they want to stay relevant, or somebody else will disintermediate them. And you'll just be a bunch of API's and a back end, yeah. But you. So that will happen. But what I think the system integrators will do is we're going to put out agentic applications. We've shipped 22 of them, and they span a broader process. They handle a bigger cognitive load. So part of the value proposition of these agentic applications, we have one that's designed to source. So I can start with a CAD design, so I have, you know, a product engineer. They look across Bill materials, they look across sourcing events and supplier data and and the system. These agents can understand context across a much broader process, right? Humans can't do that. They would be a team of humans, and I'm sharing information clever. So we're going to be able to solve broader problems. But what can happen now, system integrators can come in and, hey, look, in this industry, you know, product design is done this way. I can bring in or create a different type of specialist, and I can define what that specialist does. I can build in the expertise based on how they should look at this design or this CAD design, and identify what inventory types I need, and so they can build in that specialty into the agent, one of the agent specialists, and make it their own, or they can bring in agents from other systems and make it part of that agentic system that we've created and have it mix and share context with other agents. So they will have a significant amount of value to add it, because what we're creating is a whole kind of knowledge, knowledge infrastructure that sits under this agentic gap. Now this is how we've approached the problem. Okay, we define a problem set. It's a broad, you know, set of context. You know, the agents can reason across that and share com, you know, share information. And it's kind of these agents all working together. Others might approach the problem differently, but the end of the day, it's going to be this agents that are executing work around a defined problem set that you're trying to solve that's that's kind of how it's going to come together, Systems of Record vendors might might solve it differently, but that's essentially what they're going to have to do.

Trish  26:59  
Yeah, what kind of reaction are you getting? Because systems of outcome? I love there was one thing I read that you were talking about Hudson not task oriented. And so many of us in this world right now, we are task oriented, right? We've sort of been led down that path, right when everything went into the cloud. What is, what is that next step? Like, whether I'm in an enterprise size organization or maybe an organization that's a little bit smaller, like, what should I be thinking of, and what is my reaction to that's a pretty major, big shift, right? What they're doing?

Chris Leone  27:34  
Yeah, I mean, so I'll give you, we'll go back to the sourcing the command center, which is, you know, managing kind of end to end sourcing events. And, you know, looks at demand, it looks at inventory, it looks at a broad level of things to create the sourcing events and all these things and so. So what we can do with if we give the the agentic application, an outcome, a goal, and we say, hey, look, your goal is to reduce supplier spend by 15% and reduce lead times by 10% each one of those agents that is working when it's creating the RFQ, when it's like, you know, when it's identifying which suppliers they need to respond to, or you know, the prices that they're going to ask them to come in with, who they're going to reward winners to that do they break the auction up into different winners based on they always reason over the goal or objective of the agentic application. So each time they're recommending an action a next step, this is what you need to follow up with the supplier. This is what you need to communicate them in order to get their bids in line. It's always going back to the outcome that the agentic application is trying to achieve. So when I say it's an outcome, it would be like us creating a team, and our boss is saying, hey, look, you need to go reduce our spend by this amount. How are you going to go do it? And when you go communicate to the suppliers, it's like, hey, we need to reduce our spend by 10%. Here's the auction. You need to come in this level. It's the same thing. That's the outcome that you want to achieve. Instead of humans doing it, we're having these agents collaborate and share information, but always going back to reasoning over that that defined outcome, and those outcomes are different from every processes. So you can go and define those things, and when it recommends the next action to take, the next follow up, the next communication to do, it's always thinking back to what that outcome is. And so that's how we've thought about assembling this set of agents around a pattern where they can share, they have memory, and they can collaborate and share, context, all those things. It works from that perspective. Others can approach it differently. But ultimately, when I say it's about outcomes, it's that defined outcome that you're trying to achieve.

Steve  29:36  
Last question, I think, for me, Chris on this is, what does that mean, would you say? This could be a bit speculative, since this is really early days on these technologies. But what do you think is the major outcome going to be to managers or leaders? Right? Who maybe before I would have been a head of that team, and I would have tasked seven people on the team to go do all the seven different elements. Right of this sourcing process that you described, and now it's going to be run many, in many ways, by agents like what I'm not, I'm not talking about role going away or not. I'm not talking about that at all. But just, how do I think about it differently, and how do I sort of manage that, whether it's, you know, an AI agent on my team kind of thing. Where does that leave? How we're going to run business?

Chris Leone  30:20  
Yeah, I think that, I think there's going to be a kind of marriage of human and machine. And I don't say that. I just say that it has to happen. I think humans will always be part of the decisioning process. They will have oversight. They will have judgment. They will make ethical trade offs. They will they will monitor what's happening. And as I said earlier, they may hit the like, approve, approve, take this action. Take this action. Takes this action at the beginning, and then later on, they might only do it 50% of the time, and then later on, they might be just looking at what the, you know, the team of agents did around a particular process, and if it runs into an issue, just managing those issues and perceptions. And so I think, I think that's how work will happen. And again, it, I think it will go process by process, role by role, and you'll be able to define what we can automate in this role. And you'll take these roles that have a, you know, a large need to flex, you know, labor and, and you'll work on those first and then you'll go to the next one, you know, go next one and, and that's how I think it will roll out. And, you know, is it going to mean that? Yeah, we will mean that we do less kind of manual things. We're not going to go to a screen and type in a bunch of bunch of things. We're going to look at the RFQs that we've done for the last, you know, five years. We're going to understand what content we put into them. We're gonna look at our inventory, and it's gonna create those things, and we'll just review it and approve it. Or in some, I mean, I'll give you an example of where we can add value. And this is something we've already built, where you couldn't even do it before. There's a lot of these. I don't know why I'm on the sourcing when I sourcing one. I could go to I give you some HR examples too.

Steve  32:05  
Good example, though, because it's really, really complex. Though it's much more complex than I said earlier. Generating a job description.

Chris Leone  32:10  
I have some good HR ones too. We should do great one. But there's a number of auctions that happen that are so low value. There's that it's not worth having people manage this process. So what they do is they end up overpaying for pens and pencils or syringes or whatever, because it's too costly to go and do all these like low value auctions. Now we can do we can just completely automate that, right? So we can just, hey, go run all these auctions low value. Run 1000s of them, make sure the quality is there. We're working with certain vendors, and I can save 10, 15% on a lot of material, what that we buy in bulk, right where I never could do that in the past. I didn't have the capacity or better. So those are the types of things that we can do just that we just didn't have the bandwidth to do in the past. And so, you know, those are just new opportunities to roll out these true automations that we just couldn't, could never do.

Trish  33:05  
Yeah, well, and I think that example is a good one that you could, you could have that conversation with any C suite leader and they would understand. So I think a lot of the maybe it's rumor out there that just flows, is that, you know, leaders aren't understanding what the opportunities are with agentic AI and having big chunks of the work being done by them. But what you just said to me, if you were going into a potential customer or existing customer, right, and saying that, they're going to say, of course, I want to do that exactly. Let's do it today, right?

Chris Leone  33:37  
Exactly. So the reason, not the reason. I mean, we put agentic apps out there. It's a whole new architecture. We designed it specifically to be agent back, meaning, like, it's not a transactional system. It's like, it's a team of agents that work in, you know, against a desired outcome. And the UI is very, very lightweight. It recommends actions and communications. It's all contextual, completely different. We didn't start, you see, this is what I keep telling you. We didn't start with our transactional system and say, hey, now we're going to try to augment this. What we we've added tons of value in the flow work. So you have assistance, you have embedded AI, we have all that. But we started over and said, We're going to create these agentic applications that are completely fundamentally different. They're backed by agents. They share information, they share context. So we started over because we believe that we have to kind of reinvent how these systems should work. But we get the value of sitting on top of all of the goodness and have so you don't have to change your security model. You don't have to change your permissions. All the auditability is already there, so we know that it's going to be like, it's going to work, and things are going to get done in the right way, but we can think about this completely differently. So I think some of that is what people need to kind of start to understand and see. But when you I don't know what the end UI will be. Ours is very, very lightweight. It brings actions forward and is very, very dynamic. It's not going to be a chat interface. I don't it's not. I think the world will evolve beyond that, you know, I think, I think ultimately it might just be an exception based list, meaning, like I might be using a, you know, a claw channels, or I might be communicating with it through some different source and saying, Hey, run these sourcing events for me, and the system goes and does it I just check in on the exceptions, you know? I don't know what the end result will be, but the end of the day, you need to have a framework like agentic applications to go run those processes, define what the goals and objectives are and execute work. And if you don't have that, you're just going to be behind, you know. So that's, that's that's kind of how we.

Steve  35:42  
I think it's gonna be like, like, I'm looking for the whack a mole interface, so, like, all the problems will pop up and hit it with.

Chris Leone  35:51  
That's what it will be. Hopefully there's not that many and it will work over time.

Steve  35:52  
Hit it with a hammer and an agent takes it away, right? That'd be pretty cool.

Trish  35:56  
Yeah. Like, we've wanted this for so long, we've been talking about this for decades. That we want.

Steve  36:02  
This is a great, yeah, this was a perfect conversation for System of Records. We did talk about System of Record a lot today, but it's what there's some really fundamental changes happening in enterprise tech. It probably well, who cares if they're more important than the last bunch of big changes? They're important, right? And they're really they're really interesting. And I think, you know, you could tell you really jazzed up about it. I got super excited, like, literally doing a 17 minute monolog on this last week that I'm not sure anybody, like, really understood. But it was so much fun to do it. I think there's more to the story here. We'll probably save it for the next time we have you on maybe a little bit down the road, but we'll share the press release. I'm looking at it right now. Some apologies. Be looking away, but it's on my other screen. Oracle introduces fusion agentic applications. We'll share the link to that press release. Oracle's doing a lot getting the word out. London last week, New York this week, at an AI world event, I believe this is really compelling stuff, interesting stuff, transformative stuff, and I mean that very sincerely. I'm excited Chris and I could tell you are too. So thank you for sharing some of the story, getting behind the inside of your viral posts, right that we've been reading and enjoying lately. So thanks again, man. We really appreciate your time today.

Chris Leone  37:18  
Anytime. Great time with you both.

Trish  37:20  
Yeah, and keep up the writing. I live for those posts. 

Steve  37:23  
Your favorite I know Chris is your favorite guest. I'm glad you were able to be on System of Record. How about that?

Trish  37:29  
I know it's great. I'm like, a total fan girl. So yeah.

Steve  37:34  
Now really good stuff. Okay, go to hrhappyhour.net. Get all the show archives, links in the show notes, etc, etc. Thanks, Chris Leone, thanks to our friends at Oracle for helping us make this happen in a very busy week. My name is Steve Boese, this is System of Record. We'll see you next time, and bye for now.

Transcribed by https://otter.ai