The Self-Driving Enterprise Is Already Here with Fred Laluyaux of Aera Technology

Speaker A: New jobs being created. Decision Architect. We enable people with our community. We're now literally building training courses for Decision Architect, Decision Analyst. So a whole new set of jobs being created. That transformation is not just era. I couldn't care less. I care about the impact on the industry in my career. You know, you were kind enough to ask me to talk about it. I've seen, you know, mainframe, client, server, cloud, blah, blah, blah. This is different because none of these technologies has fundamentally changed the way people work. This one fundamentally changes the way people work.

Speaker B: M welcome to CIO Classified, where you'll find candid conversations with the world's leading CIOs. This week we were joined by Fred Laliu, co founder, president and CEO of Aera Technology. In this episode, Fred breaks down what Decision Intelligence actually is, why data was never the real bottleneck, and what it takes to get AI agents making real decisions at scale, not just recommendations. Now here are your hosts, Yousef Khan and Ian Faison.

Speaker C: Fred Laloux, welcome. Good to see you. Bonjour, sir. Uh, yes, it's been too long. Wonderful to see you again.

Speaker A: How are you?

Speaker C: I'm, I'm doing great. I'm doing great, I think. I, I. It's been quite the journey for you and I've learned a ton. Um, so we're going to get straight into it. Welcome to CIO Classified. There's a lot of CIOs who are listening to this. They want to get a bunch of advice. They probably want to drown their sorrows as they listen to us in their car driving up the 280, trying to figure out where life takes them, especially with all the stuff that's going on. AI. So would be great to first start. Um, you and I have known each other for a decade. Uh, you've been at least, yes. You've been an operator, taken a company, taken multiple companies now to the next level, taking parts to market. And, uh, I've really admired your journey. I've learned a ton in the interactions I had. So I look forward to reconnecting with you, not just today, but also after this. So let's just, let's just start off a little bit because, uh, when we last, our last time we spoke, uh, you had just started, uh, at era, and prior to that, of course, we had interactions. You had taken out a plan, just, uh, to the sky and then some. Um, first of all, how are you doing? And, uh, would love for, I think a lot of people in the audience would like to actually hear about a little bit about the career Journey. Because a lot of CIOs who listen to this always think about the next chapter. They get to a operating level and they want to think about where they go next. And you've, you've done that in your career through SAP and others. So we'd love to just get a little bit more background context and then let's dive in a little bit more about, about all the activity that's happening in AI.

Speaker A: Sure, sure. Well, um, where do I start? Uh, born and raised in France, uh, entrepreneur from the very beginnings, uh, jumped into the enterprise software world in uh, 1997. Spent a few years building a company doing cost modeling, cost management, performance management, building the category around enterprise performance management. Moved to the US in 2002, uh, spent a few years, we sold that company to Business Objects back in the days, um, you know, that was in 2000, I don't know, 2006 I believe. And then PubJ got acquired by SAP. Spent a few years with SAP. I was in charge of the post merger integration between SAP and Business Object and I did some strategy work. Uh, then I was in charge of the finance and analytics line of business for SAP. That was around 2010. And this is when the idea of era really came up. Left SAP, built Anaplan, uh, for a few years, did good success there and the idea that I was uh, nurturing in my mind was still becoming at that time more and more uh, obvious that there was the next big unlock in enterprise software in enterprise performance management and sustainability. So I decided to move on and uh, we built era and we launched era 6-19-2017. Which is crazy. Nine years, uh, hard working on that adventure, uh to this day. So that's the quick version of my career. Um, but I've been an operator and an entrepreneur pretty much all my career.

Speaker C: It's amazing. So let's get into it a little bit. Clearly you know you started in 2017, um, and I remember like even reading about it when you were talking about decision intelligence then for us, you know, I was uh, looking at Moveworks at the time from an ITSM and automation from that standpoint. But I think it's very clear that you've always dealt with uh, well I would say big data pretty much from the beginning of your career. I mean. Yep, Business objects and a plan. Now to Aera. So I guess, I guess the biggest question is tell us a little bit about uh, AERA in terms of how you are being able to consume massive levels of data, being able to actually understand how a business works and then Helping to be able to drive decision. Would love to just hear about, about the platform a little bit and then let's, let's go.

Speaker A: I think it might be important to calibrate why, why the journey and why they decide to do to launch this company. And you talked about 2017 by the way. At the time Decision Intelligence was not a. When we launched in 2017 we said welcome to the self driving enterprise. That was our paper that we used to, to launch the company and announce the company. Um, so why, what was that vision between for the self driving enterprise and for me it was anchored on a very simple, you know, belief I should say or conviction that the current model where companies are you know, leveraging uh bedrock of transactional systems and on top of which they build a bunch of data products and then they have a bunch of systems of differentiation applications and then they have a layer of analytics on top to try to disseminate the information across the enterprise. That model was designed and still is to enable people to manually make decisions. And I realized that back in 2010 when I was at SAP that for me the biggest unlock that we could work on and it's becoming an uh, obvious thing today. But back then it was not fundamentally was that the digitization of our economy overall was creating three things. An acceleration of our business cycles. Back in the anaplan days, our main reason raison debt was the ability to plan faster. And when I think about it today, it's a joke, right? We were like okay.

Speaker C: By the way, if I may interrupt Fred, one of the things that I always admired and respected it is that you ran Anaplan using Anaplan like you

Speaker A: uh, today we're running error on error.

Speaker C: It was uh, you always driven that I want to make sure that people understand like it's not just oh, is a product et cetera. You, oh no, from a customer vantage

Speaker A: people say here you, you eat your own dog food. I like to say we drink our own champagne, but that's my French. That's the French in me. So back to the point. Right? So three things that were really driving the idea behind ERA uh, it's faster just rhythm, not just decision cycle, but faster rhythm in your enterprise. The second is that the consumerization of our enterprises driving decisions that you have to make at a much finer level of detail, fine grain. And the third is the complexity, uh, that is impacting our company. So my simple starting point was if I combine the speed, the volume and the complexity, I'm going to get to a point where that model that is structured and designed for people to make decisions will not hold. And the paradigm shift that we've announced when we launched a company was we want to move from people making and executing decisions supported by machines, data sets, collaboration platform, bespoke tools, to a world where machines can make and execute decisions guided by people. Hence the analogy with the self driving car and talking about a self driving enterprise. So fundamentally the starting point was like how do I digitize decisions in a large enterprise? And along with that was a second realization after. At that point I'd spent 20 years, uh, uh, doing performance management performance improvement solutions with the world's largest companies. Was that um, and I still have that questions when I talk to CIOs today. You have all these great platforms and data lakes and this and that and cloud. Where do you store the decisions? Where do you memorize the decisions that you make thousands of times a month to, to improve the and run your business? Where is that sitting? And they look at me going like wait, no, it's in people's head. So we're in 2026 Yusuf. And you think about performance improvement, you think about AI, uh, you think about all of this. And the reality is that our models have not changed for 100 years. We're still relying on people to decide how to forecast, how to allocate inventory, how to change a plan. Whereas if you think about a world where you digitize those decisions, you can memorize them and that memory becomes the foundation of continuous improvement. And that was to me a starting point, like I got to find a way to build a solution to the market that allows large complex organizations to memorize the decisions that they make. And, and by doing so, that will be the foundation of continuous improvement. That will be the foundation of thinking and operating across the value chain. That's why we started era, but we started from the bottom up. It's from crotch.

Speaker C: I think it's really important to point this out. If I think about your career, you've always been in the decisions business.

Speaker A: Yeah.

Speaker C: I mean if you think about what you've been building business objects, then anaplan and now era, like you've been trying to get closer to be able to make businesses making those faster decisions. Correct. I think what's clear to me is that ah, the transition from the old world, if I could call it that, from taking analytics and dashboards and being able to ingest data to now being able to drive enterprise and decisions at scale is kind of where the biggest opportunity is for Aera. I guess m my Big question is, if you think about now, three and a half years post ChatGPT, let's think about the AI reality check. And so CIS are still under pressure on the ROI side of things. So why has it been so hard? I mean, you know, you've got more ability, you've got more engineers available to you to be able to build something. The models are getting better, that something's not m, something's missing. What are you hearing from CIOs? What advice do you have for them to be able to sort of move forward by um, getting an ROI initiative?

Speaker A: Look, to this day I think ERA has digitized more than 50 million decisions for the likes of Exxon, Unilever, Estee. I mean the largest companies in the world. What's different between what they do and what others do? They look at a leveraging the capabilities that you just talked about, agentic LLMs and others to solve the problem at the enterprise level and not to provide. It's the paradigm shift. And not to provide just a yet another better personal productivity solution. Copilot is fantastic, but it's a personal productivity solution. The paradigm that is shifting today is really thinking about this technology to work autonomously on your enterprise data. And I know that's uh, something that we need to unpack as opposed to trying to get the human beings working in the company more efficient with this technology. The paradigm shift is happening. Right? Uh, Gartner comes up, released this uh, January, the first decision intelligence Magic Quadrant, right. And they're talking about a couple of very important points. Actually there are three points that we always like to highlight. It's the shift from data driven MHM to decision centric. You have the shift from I'm trying to provide data, data products and data solutions for my people to I'm starting with how do I make decisions. And that's my starting point and that's my ending point. How do I digitize a decision that I make? So the shift from data driven to decision centricity is super important. And they also project a couple of uh, elements. 50% of enterprise decisions will be touched by uh, decision intelligence by 2027. And we're seeing that movement right now. And again it's not just about one technology. It's decision intelligence covers multiple uh, capability. And the last element, which I think we're seeing right now, and we've seen it through our journey when we started and we deployed the technology to some of our uh, clients, the initial reaction, the technology delivers real time recommendations, it will look at a forecast, it'll say there is a stock at risk here. I recommend that you rebalance inventory from this DC to that DC in the next two minutes. There is a, not two minutes, whatever time frame is. There is a uh, change in the weather. I recommend that you move your stock from here to all these recommendations. Initially they were provided in an inbox and then companies were saying we need our operators to review them and accept them or reject them. And then the system goes and executes the decision back into your transactional system. The shift here is that we've got enough data, I mentioned the 50 million decisions to demonstrate that whenever the humans are touching the system and are messing with the recommendation they actually degrade the performance. Not all the time, but in aggregate it does. So we're seeing the confidence in the machine driven recommendations, decisions and actions taking over whether and then you use the humans to guide the system. And if you allow me, the funny thing is, and the interesting thing, not so funny actually is the analogy with the self driving car. Since the moment we started we said welcome to the self driving enterprise. We put a picture of an enterprise and a waymo, uh, at the time which wasn't yet allowed to drive on the road. And the evolution of DI with self driving uh, uh, technology has been, the parallel has been amazing and today a lot of people say I'd rather put my kids uh, uh, in a self driving car than in an Uber because I don't have the risk of the driver doing something silly. So the confidence is now shifted towards self driving technology as opposed to where it was just a few years ago. The same thing is happening with enterprise decisioning where the confidence in the quality and the accuracy and the timeliness of the decisions made by era is increasing over the need to actually review every recommendation delivered by the system.

Speaker C: Um, so I'm going to just uh, hand over to my colleague Ian. Uh, but I do think I want to make sure that something's clear. The decisions that are being made uh, at the companies that you've mentioned, the customers, I mean to get this right, is affecting tens of millions of dollars of revenue. And, and I think just, I think something, sometimes people just miss out the level of scale that provides. So, so Ian, but Ian had some interesting points to, to raise us.

Speaker D: Ian, so when you're talking about sort of like digitizing decisions, like what does that actually mean in practice? Um, because I think that every single like when you're you know speaking about uh, the complexity of the enterprise and how decisions are getting made and how many of those don't, don't need to be made by a human. I think we all like sort of nod along and agree. But what does that look like on a day to day basis?

Speaker A: Yeah, I'll get to that point in uh, a second. Just to Yusuf's last comment. Um, I was quite pleased to see a couple of weeks ago, um, the Hershey Company, a charter manufacturer and um, in their um, uh, analyst day. Right. Financial analysts have on their presentation a uh, slide that talked about how Decision Intelligence was going to save the company $50 million and reduce inventory level by a million dollar by $100 million within the next couple of years. So the impact of what we're talking about is very real. Now Ian, back to your question. I, um, think somehow beyond the question you're asking me, how does it work? Right. I think that's maybe what I can describe. So if you want to digitize a uh, decision. So think about a decision is what an operator does in your business every day. They're managing logistics, they're managing procurement, they're managing finance, they're managing promotions and pricing. They are making tens of decisions every day. How do I create a promotion that makes sure that my inventory gets consumed before it gets obsolete? Right. That's an example. How do I forecast, how do I locate things across my factory, how do I manage my contract manufacturers? Those are the very pragmatic day to day decisions that you have to make. So how do you make the decisions? Either there is an event that, that, that you need to take care of or it's on a schedule. Every Tuesday, every morning. I need to do this every day. I need to look at how I'm going to uh, um, allocate my semi finished, good, from this factory to that packing center and which one. That's the decision that you have to make. And to do that, to make that decisions, you have access to data, you have access to software to help you run some kind of calculation. You're going to project, you're going to predict, you're going to optimize, you're going to allocate, you're going to decide, you're going to simulate, that's what you do. And you're using bespoke solutions for that and the outcome of that work and you're going to comply to policies, you're going to follow specific guidance, you're going to be compliant with processes and so on and so forth. So that's the work that people, millions of people do every day. If you're a planner, that's what you do and that work, the outcome, the work product that comes out of this is a decision. And that decision, you might be able to make it on your own or you might have to go and talk to somebody, or you might to inform a bunch of people in your network and say, hey, I'm going to do this, am I allowed to? And people are going to challenge your level of confidence. They're going to ask you questions, Ian, why are you doing this? Have you thought about this? So there is a process here and the outcome is uh, so you come up with the recommendation, the outcome is a decision and then the next thing you have to do is go and execute the decision. You're going to create a stock transfer order in SAP, you're going to change your planning parameter, you're going to look um, at your master data and ah, make an adjustment and so then that decision is executed. Well, what I've described is exactly what Aera does. ERA connects to your enterprise data. We have technology patents, we're doing it with the world's largest company from Nvidia to Dell to, to Unilever to Exxon to you know, largest companies in the world. So connect to their transactional system, external data sources structured and structured and bring the necessary data to uh, build what was called a decision data layer, the data model. That data model has your metrics that it calculates right. What is my on time in full, what is my lead time, what is my blah blah, blah. So that model is the foundation and it's got something quite unique because it's got the real time representation of the complexity of your business. It understands it, but it uh, will also store the memory of all the decisions that are being made across your enterprise. The only interactions that you have with the tool and those decisions are made in uh, another layer which is comprising of two dimensions. The first one is what we call agentic ambient uh, intelligence, which is the reasoning part of the system which then can be both deterministic or agentic or hybrid which is, I'll come back to that because that's super important. But this reasoning needs then to work with a set of engines in the platform that we've built over time you talked about on our plan we basically rebuilt that same technology much better. But we also build a machine learning, we build statistical forecasting, we build self learning, uh, engines with a bunch of engines graph and so on and so forth that allow you to do the math. So Ian, if you go back to, I know what my process is now I need to trunk Process the numbers and come up with the outcome. So it's the decision data model, the multi engine orchestration, the agentic ambient intelligence. We'll come back to that point because it's super important. And then it's a layer of engagement. I need to engage with people internally, externally, I need to engage with data, I need to engage with systems as I write back the outcome of my decision and I need to engage to agents because our customers today have built agents, they have a bunch of agents everywhere. So how do I leverage some of that intelligence to improve the quality of the decisions that we make? That's how it works and it's infinitely scalable. So a lot of companies today, if you think about how they work, they have classification everywhere. So why do you classify? Fundamentally you classify because you're trying to assign your limited amount of human based compute power to the decisions or the problems that are the most critical. ERA doesn't have that problem. The logic that I deploy can now be applied in real time across my entire bill of materials, my entire set of SKUs, my entire set of consumers. And this is where you get the value unlock that I talked about. Combined with the fact that we creating that memory that doesn't lie. I can now start. This is why it's called decision intelligence. There is a layer of intelligence that you apply to that memory and it doesn't lie. It say that when we make this kind of reappointing of sales order decisions from here to there over time, this is the impact across my different business dimension, cash cost, service level, sustainability, water, power, whatever it is. So you have now this new gem of memory which is very easy to build. We can deploy ERA record speed today and you start building that memory. And that memory doesn't lie. If someone leaves their job, if your best planner leaves the job or takes another uh, job in the company, the best practices, the memory of that person, the way they did the work doesn't leave the function. You see what I mean? So that's really the unlock that we're creating here.

Speaker C: A lot of stuff is happening in the enterprise as a result of deploying better decision intelligence. I think one of the things that always comes up in CIO conversations is about the level of trust. I'm not talking about, you know, Claude and uh, you know, source code and all the other stuff. I'm just talking about core basics, which is about governance and trust and how CIOs need to be able to build that. How do you think about that from an autonomy standpoint? Let me Just paint a picture to you. A lot of my friends have not been wanting to go into Waymo. I love Waymo. Okay. It's phenomenal. It's, you can take a nap, you can take a zoom call, you can put on your Spotify playlist. It's fantastic. I guess the question to you, Fred, is how do people get into a place where governance and trust is put into place when it comes to AI driving decisions and the companies that you are serving clients. This is tens, hundreds of millions of dollars of revenue. Yeah, right.

Speaker A: Billions. Yeah.

Speaker C: And there's change management required for somebody to say, oh, we don't need to have a meeting about this, this is just being done. So could you just talk a little about the, maybe the obstacle of being able to get people to move in this direction that companies need to think about. CIOs want to feel that they need to be able to have this conversation and they always struggle to it. So what could you provide to be able to give a little bit more depth in that?

Speaker A: Well, there are two levels in my answer. The tactical one is trust in the software. Uh, being one click away from being able to understand the data sources, the logic, the algorithm running, the quality of the memory. We provide that control room that gives you full transparency on what is happening in the platform. Put that in perspective, Yusuf, with the fact that today you have no clue how decisions are being made. So that's an argument. You know, I talked about the memory earlier. I said, where do you store the memory of the decision that you make, which is such a logical thing to do when you're trying to improve performance over time. But if you don't record and it's ah, a ground out there every time you make a decision, we're providing something here that gives you a more trust because now you understand what you do today. You think about a company with 80,000 employees that is operating in 100 countries. Do you think operators, uh, and CIOs understand how decisions are actually being made?

Speaker C: May they don't.

Speaker A: They absolutely don't. It's, it's, it's, it's not controlled. If we going back to the end to the WH analogy, I think yeah, you should ask them to sponsor your podcast now. But uh, if you go back to the WIMO analogy, they understand how every car operates on the road. They have the data. Do you think Ford or GM or Mercedes understand that? No, they don't. So trust. Let's go back to the basic. To trust you need to understand and today you do not Understand, by the way, side note, but interesting. When we started, we engaged on pilots with uh, with great companies, always pretty significant size, very advanced. Uh, so I'm not talking about a startup here, I'm not talking about. We went with the Fortune 100 and this is where we deployed the technology starting at the beginning. And I did that for a reason. We can come back to that. Um, and when you digitize decision, the first thing you have to do is tell the system how you want the decisions to be made. You're going to configure ERA to say if there is, you do a forecast, if you identify a gap. This is the decision that you can make. You can do this, this, this and that. And this is how you stack, render choices and blah, blah, blah, blah, blah. Well, guess what? They build it self service to build in the platform. Then they start rolling it out. And what did they realize that at the, at the end point, the operators 6 level below the CIO who are making those decisions were like, hey guys, what, what is this? What is that? This is not how we work. And you go like, well, that's how you supposed to work as part of the uh, this is how uh, we think you should be working. So I'm talking five, six, seven years ago. But that was the first unlock, which was you. And I've said that a million times. You do not understand how decisions are being made in the company. And when I talk about decisions, you're thinking strategy, you're thinking, should I open a new factory, should I stop? But decisions are everywhere. It's the millions of structured decisions that you're making every day. So we build these capabilities in the platform. We've, you know, now it's less of an issue. But the first argument is, do I trust AI versus my people? Okay, let's have a conversation. AI will give you, ERA will give you a undisputable memory of all the decisions and analytic and intelligence. We have a client who uh, as a result of having deployed AI in multiple areas, has been able to reshape some of their policies. The system recommends policy changes and say, look, your operators are rejecting the recommendation. And they're right because this is actually stupid or this needs to be improved. So Yusuf, the conversation about trust is comparative right now for me, uh, just like probably the folks at wimo and Zoox and others, I cannot afford a single accident, right? So there is one big topic which is when you think about AI, you think about era, you think about agentic.

Speaker C: Great.

Speaker A: We are one of the first technology out there to fundamentally leverage the full and harvest the full power of the agentic reasoning in our platform. But when I'm going to trust ERA to run sales order reporting uh, uh, um, decisions a thousand times a month, I don't necessarily want that to be agentic. I want that to be deterministic. I want that to be a software with code and lines and boxes. So what Aera does is it allows you to leverage the dynamic reasoning to come up with the solution of the agent. But then ask Aera to build a skill, a competency, a set of decisions that you're going to digitize, you're going to instruct it to do it maybe in a deterministic manner and it will go and build a skill in some areas. Let's say I need to read a claim that's coming from a PDF or I need to look at a picture of inventory in a warehouse. Whatever it is, I will still need agents but I need a very narrow scope of agents. So uh, that's, I know it can be controlled and supervised but the full agentic reasoning is mostly leveraged to identify the opportunities and to help build a deterministic skill. And I think that's going to be critically important because I can imagine that hyper relying on LLMs will becoming a problem when uh, those LLMs are less available because power because reason X, Y and Z. We had an issue a few months back where um, Microsoft, uh, Azure was not giving us the nodes that we needed to run the system. Wow, what a wake up call my friend. What's going on? Where is my capacity? Because ERA consumes a lot of capacity and uh, for a couple of weeks they had a snitch, they had a problem, they couldn't solve it. Think about now if you're relying on LLMs to do the decisioning in your company and this happens, your inventory doesn't move, your factory don't get the raw material, your forecasts get whacked. I mean this is really important to know and build the trust by explaining that the hybrid mode between deterministic logic agentic, uh, is, is available because I wouldn't want to run uh, you know everything on agentic.

Speaker C: Let me ask you one thing uh, Fred, and we've just got two uh, more questions. The first question is if you think about the deployment of this in at large, can you just talk a little bit about the technology stack like what is a good customer for ERA where you know, you just need to you know, is it replacement, you know, is it orchestration? Is it both? You Know, where do you fit in really well, like if, if that's hypothetically assuming.

Speaker A: Yeah, great question.

Speaker C: I was on if I was on the road, you know, as an SDR for era, and I was saying, look, you got to meet my friend Fred and get all my CIO friends together where what's a great fit for you from a customer that you should, that should be adopting in the technology stack terms of where you fit. So.

Speaker A: I'm going to give you an answer that's going to sound not very credible, but it is the answer. We don't care. Um, we love SAP, we love large erps. That and especially if you have a messy landscape with plenty of them and you've never been able to harmonize your data. Uh, we have a client in Germany that harmonizes I think 42 instances of transactional systems into AI.

Speaker C: Make a suggestion. I think you should call it Harmonize AI because I think that's where we're getting to. It's not an orchestration layer, it's not replacement layer, it's a harmonization.

Speaker A: Yeah, but that's just the first part. That, that's just the first part. We had to resolve this problem. And my friend, co founder Sharik, that's what he basically did when I met with him. M. He had solved that problem and it's, it's, it's deployed at the largest companies in the world. Um, so being able to connect structure and structure, I don't care. Dell doesn't use SAP. They use uh, the homegrown application with some other stuff. Just to put things in perspective, maybe I, I should mention that I talked about Hershey's earlier and they were on stage with us at several confere from the moment they started the project to the moment they were live with. Three skills. Delivering recommendations, taking action. Three months start to having the system in production from that point on. And it's publicly, they've talked about it publicly from that point on, it took another 90 days to get a full ROI on the project. Why? Because they can measure the impact of every single decision. An improvement made by the system. We're getting to a point where we're going to announce in the next few weeks, um, uh, what we call internally the 444project. Four minutes to sign in to Era Connect, load your data, four hours to get. Wow. Because now you're interacting with the system and it's thinking through your data with this industry memory and then four days to have a skill in production by mid year. We're at that level. Okay. So now where does it sit? There is a concept that I talk a lot about called the collapsing of the stack. If you think about it, you're a cio. I met you, you were a cio. Uh, you have your transactional system as I explained before, and then you build data products everywhere and you try to connect them with analytics and copilots and agentic platform and AI models and you try to weave that together. If you're thinking about planning, you're going to have your system and a planning tool and this and that. The collapsing of the stack is happening. We can work straight on top of the erp. And now if you have a bunch of these products, we work with them of course. But the stack of the future is going to be a system of dynamic intelligence like era and a system of record which going to get smarter over time. And the reason I say that, going back one more time, maybe the last time. On the self driving analogy, a uh, Waymo makes no sense. It's a fully equipped car for a driver to operate.

Speaker D: Right?

Speaker A: Right. It's designed as a car, it's got a steering wheel, it's got a lot of dials, it's got all the stuff that allow a person to drive. And on top you stuck a bunch of raiders, lighters, sensors and intelligence to make it drive autonomously. And then you have the Zoox, which is the Amazon product that's uh, driving in the streets of San Francisco. The next generation. You remove the human capability. The human cannot drive a Zook. It's basically a platform on wheels. And I've removed all the stuff that I enabled uh, to make a human. The driver. Think about the cost, think about the effectiveness, think about, about everything that goes with it. The way I think about it is our enterprises are full of people who are driving manually. So the work product of a Zoox is the same as the work product of a wimo and is the same as the work product of a regular car. It's allowing me to go from point A to point B, but I'm not going to need steering wheels anymore, I'm not going to need dials to tell me the speed and stuff, uh, anymore. I'm not going to need gear, uh, stick, I'm not going to need any of that anymore. And just apply that to your software stack my friend. This is what's going to happen. The Autonomy is not another version or better version of my planning tool or my replenishment tool. It's not that it replaces the need to have a human touch that Software and therefore I don't need that software anymore. I bring the logic that's being deployed, that's being enabled via this software. Talked about on a plan. I can do a plan. Well, I can do a plan in ERA without having people keying things on the queue. If I need to, I can, right? But the stack over time is going to collapse. You're going to have a system of intelligence system and a bunch of uh, systems of records. And you have to be able to put your hands back on the wheel in your system of intelligence if you need to. But the system of intelligence will allow you to control, govern and drive the machines that are driving themselves. And that is part of our vision for building era. And this is starting to happen. We have a customer in New York, fantastic company, growing like a weed and as they're growing so fast to say we're going to build a stack and their stack is SAP and ERA and nothing in between. When we deploy into an existing organization, we connect to the transactional system, all the systems of differentiation that they have. We might be pulling data, uh, in and out of their databricks or their snowflake or whatever that is. That's fine. We adapt and it's the layer that connects the dots. The last point I will make on this, which is absolutely fundamental. I, uh, talked about the collapsing of the stack. The next level is the collapsing of the organizations, right? The delayerization of organizations will be enabled by these tools as they think across the value chain. Guess what? We're already seeing that in New York in uh, uh, actually October, I think we have our user, uh, conference. You'll hear customer talking about the impact of the technology on their stack, on their organizational design and on their talent. New jobs being created. Decision Architect, we enable people with our community. We're now literally building training courses for Decision Architect, Decision Analyst. So a whole new set of jobs being created. That transformation is not just era. I couldn't care less. I care about the impact on the industry. I think a lot of vendors are going to be a bit worried and they should be because at this point in time we're not competing building a better version of the existing tools by, in my career, you know, you were kind enough to ask me to talk about it. I've seen, you know, mainframe, client, server, cloud, blah, blah, blah. This is different because none of this technology has fundamentally changed the way people work. This one fundamentally changes the way people work.

Speaker D: I think that that's what's really exciting, you know, about what you're Talking about is every single company inherently knows that that's the case. Right. It goes down to the household, right? It's like we have enough data of how much almond milk that my family, you know, buys that like, we don't need, um, we don't need, you know, you could set that on autopilot, right? But that's sort of like saying, oh, I'm going to, I'm going to, uh, you know, let's, let's buy almond milk every, every month and let's buy like three things. And my wife's like, I'm not comfortable with that. Whereas, um, the actual AI saying like, hey, you're running low on almond milk, I'm just going to order it for you. Like, those are the things where like, we all want to get there. And what, from what you're saying that people, you know, in three months are, you know, deploying real skills is so exciting because like, those are the, those are low leverage decisions or they're low, they're for a human, they're low leverage decisions that don't need to get made. So you can focus on the way harder decisions that are going to have much more sort of like material impact. And I love the way of just framing the entire problem through the decision making framework. Um, and I, and I use framework because it's like we have all these literal decision making frameworks, right? So if we have a framework, we could give that to a robot because they could execute a framework, right?

Speaker A: Yeah.

Speaker D: Like, that's the whole point is like give it the instructions and say, oh, well, you know, I don't need a person to go through this framework. I already have it built. Here you go.

Speaker A: But the, um, if you stick to your almond milk example, um, one of our customers is running a demand forecast at the store, SKU level store and SKU stock keeping unit on a daily basis based on the point of sale data coming from the store. So you go, so what? You know, why do I care about this? Well, because there is an entire value chain that allows that almond milk to be in your store or actually delivering to your doorstep. So when I talked, when we started the conversation about the consumerization of our economy, I mean those concepts were unthinkable 10 years back and now it's becoming the norm. There is a skill that we built with one of our partner that looks at, it's called marketing supply synchronization. It will look at the, yeah, you will receive some ad on social media for almond milk when the system thinks that you need to get uh, a new almond milk. Well, the reality is today those processes are operating in silos. Someone is looking at Ian needs to buy this. So I'm going to put an ad, that's a bit of a silly example on the other end someone is placing an order for the boxes that you're going to get them. If you could just measure in real time the current performance of that digital media campaign by outlet, by square, uh, mile, which is the pixel on a Google map, and you could correlate that to the availability of the box of milk that you like to buy in the store, then you can start making micro decisions and say, stop the campaign, I'm going to run out of stock or I've got overstock. Promote more. This is how the world works today. And those decisions, back to your question, Yusuf on trust. Those decisions, they cannot be made by humans. The decision that I called born in digital. So if you think about a company, strategic decisions, situational decisions, incident response, crisis management, structured decisions that we're currently making and we could then digitize them. But there is a whole new array of decisions that are born in digital that could not be resolved by people because the value of the decision is real time. And if you cannot make those adjustments in real time, you waste money or you create uh, service level issues as on that example. So it's not a nice to have what I'm talking about. It's a must have. The challenge about decision is the first thing that pops in your brain is well, it's strategy, it's, it's blah, blah, blah. No, no, we do that too. We cover the whole spectrum of that. But, but think about the millions of decisions that you make in a, in a business every day to just run the operation. And the way you solve this problem is you plan. I'm a planning guy. I build planning solution built on a plan. I know planning, but planning is not the answer, it's execution. If you can operate, go back to the self driving car probably since the last time, micro adjustments in real time. You don't have to plan that much. Who plans the route? Go from here to Santa Rosa, California. You don't plan anymore. You just tell you it's where I want to go. And the system automatically will make recommendations in real time and adjust to the speed of your car. And you drive Ilsttein and that's what we're talking about. It's that simple.

Speaker D: That's beautiful because I hate planning and I'm not a planner.

Speaker A: Everybody hates planning except planners.

Speaker D: I love that framing that we're, uh, not making plans anymore. It's like we're adjusting in real time constantly. Which sounds very fatiguing, but sort of if you're having a hybrid approach to it, where it's like, yeah, if you were to plan every micro decision to go from here to Santa Rosa in real time, and you had to stop the car and get out every mile that you drove, that sounds, you know, awful, but like, you know, whatever, Google Maps or whatever is just going to route you in the best direction that it thinks. I think the one thing that people perhaps might be afraid of is that, um, in using that same analogy, is that, you know, there's an accident, uh, you know, up on. On the road ahead, and it says, hey, uh, you know, Ian, you should go up Kirker Pass, uh, you know, rather than, you know, going through, uh, whatever road. And the idea that there's some other sort of governing force that's, hey, maybe once upon a time I said, I don't want to go over, you know, under freeway overpasses. But I, you know, that was three. That was. Fred did that, and then before that, Yusuf did it, and before that, Bob did it. And like, I don't even know, you know, what that, uh, that framing is. So there's that fear there. But I'd imagine that you could just query it and say, hey, what are my rules? Like, you know, right.

Speaker A: The system will. The system will say, okay, I recommend that you rebalance this to that, or you ship this there or you buy this. So you stop that purchase order, which is a very. If you're a buyer and the system makes a recommendation to put a hold on a po, you go, oh, wow, what's going to happen to me? So the first thing back to you question on trust. Explain to me the reasoning in real time in natural language. Allow me to push the reasoning with you.

Speaker C: Era.

Speaker A: That's part number one. Part number two, if you've configured the system that way, you will be allowed to say, ian, no, I reject the recommendation. I don't want to penalize this vendor because their performance metrics are going down. I want to just do a little hit. Okay? That's the decision that you make. But then ERA will ask you push that. Hey, can you explain the rationale behind that decision? Instead of rebalancing your orders from this vendor to that vendor, uh, at 20%, you only said five. Then you can enter a reason code, or you're going to write a little text and you say, I don't want to hit these Guys, the first time, what happens at that point is your information, which is the analogy with going under the bridge. Your information, your reason gets captured, memorized and manifested in the next time the decision is going to be made. And that's the governance. That's why we talk about people guiding the system. Someone is going to say, hey, um, maybe we should change the policy. And that's what we able to create is the manifestation of policy changes and it's happening naturally, which today is not happening today you don't have that. It's going to go very slow process. I think we need to make a change. Let's set up a meeting, a call quarterly. And then you argue biases come in the game. Maybe the system will look at when you're doing 5% only penalty, it actually worked and that performance improved. But now you can see it, now you can understand it and it doesn't lie.

Speaker D: Yeah, I think that this is like such a fundamental shift in like the way we work. Where you go from I used to be the guy who orders the almond milks, like I'm an almond milk buyer or whatever, you know, in this scenario to I'm managing a system every day and the system is the thing I'm managing. You know, you see this a lot of people that are working with agents where they're like, I used to be a customer success, you know, manager and now I'm managing uh, this, you know, suite of AI ah agents. But now I'm just constantly working with an agent every day to optimize all these things, which is like completely different than what I used to do in terms of tasks. But it's the exact same uh, outcomes for your customers in this case.

Speaker A: So it's like, and just think about it as ERA being the decision intelligence that not just looks at your personal work and your personal decisions, but ERA looks at the entire data set in real time.

Speaker D: Right.

Speaker A: And what we're releasing right now is mind boggling because it does feel like a chatgpt or a cloud experience in a way, but it's actually looking at uh, the entire memory, the entire context of your enterprise. They'll say, I don't think we should do this because the impact will be this. And I check the weather. I mean this is not sci fi, this is being deployed today. But it's not, I don't want people to think that it's like just making me more efficient. It's just making the decision intelligence network more efficient and I'm part of guiding that network.

Speaker C: Okay, well um, we've introduced almond milk. We've definitely talked about Waymo and we'll

Speaker A: and I think I'm going to have a glass of almond making a way more now. I think that's what I need to do.

Speaker C: I think that's the gun we're going to send the selfie to, uh, Ian on the back of that. But Fred, first and foremost, it's been really good to see you again. I look forward to seeing you in person very, very soon. Thanks for making the time. Super excited for what are is bringing to, you know, enterprises all over the world and we're looking forward to, you know, rooting you on and wishing you nothing but success in time ahead. So thanks for making time for us and a lot of valuable insights for CIOs and the listeners today.

Speaker A: It's been fun. Thanks for the opportunity guys. Really enjoyed it.

Speaker D: This episode is brought to you by ERA Technology, the leader in agentic decision intelligence Enterprise AI has hit its stride across industries. Companies are moving beyond pilots improves the concept into real enterprise wide results, better decisions, faster execution and meaningful bottom line impact eras. Agentic decision intelligence is built to help you seize that opportunity. AERA dynamically composes decision flows using unified decision data and multi engine orchestration to drive action at scale. It continuously senses what's happening across your enterprise, recommends and executes the best course of action within your transaction systems and learns from every outcome to keep improving. Leading global companies are already using AERA across supply chain, inventory, logistics and finance, delivering rapid ROI through reduced costs, lower working capital and better customer outcomes. This is the self driving enterprise and it's here now. Visit aerotechnology.com to book a demo a e r a technology dot com.