The autonomous enterprise. The beginning of better.
What makes it better? Joule. Can Joule figure out why these accounts don't reconcile? Reconciled. Also, three items need your attention. Oh, that is better. Joule. And you'll tell me which of my orders are at risk. Yes it can. And fix them. Done. Perfect. Have a great flight.
Could you tell me what happens if we run out of valve seals? Yeah. That control or more. Order ready? Click to approve. What's better? Can I find someone who's working on a product rollout in Mexico? Certainly. Is your guy. And speak Spanish. Finding new talent. Español? Sí. Yeah. She's better. Sorry. Yogurt guy.
We've been targeting the wrong segment for our best selling products since March. And you hold me? No. But it can make your targeting better. Let's do that first. Can it review this contract to flag anything I should worry about again. And send the changes before the energy arrives. How are you doing? Revised contract ready for review. Oh, that was a tie.
Can it tell me if anything is going to put us over budget? Even the ostrich feathers. I guess. Obsessed. Wait. My flight was canceled. 11 minutes ago. I found you another flight. Oh. That's better. Can you find it? Yes, it can. Better team for my kid. Are you sure it's the team? Love you buddy. You're crushing out there.
Joule, update me on our suppliers. Are they going to make the delivery? And what's it going to cost if they don't? Can you give me three options? Show me how a cash flow is. Prioritize visibility. Absolutely done. It doesn't get better than that.
Dad. Always your dad. Dad. Please welcome Christian Klein, chief executive officer of SAP.
You know what? The video. Absolutely nailed it. It really doesn't get better than this one. In the keynote in front of such an amazing audience. Welcome, everyone to SAP Connect. It's great to be back in Vegas. Just five months ago at Sapphire, it. It feels like years ago. Yeah. We introduced the autonomous enterprise. We called it the beginning of better.
And to really make it better for you, our engineers, I feel so sorry for them. Have had a very, very busy summer. Their incredible work paved the way to show you today. The autonomous enterprise in action, underpinned by great customer service. So let's get started.
You might have seen. I hope you have the opening keynote at Sapphire. I showed a drawing of a unicorn. It was fully generated by a frontier AI model. It looked pretty good. But my daughter immediately realized that this unicorn has three ears. The point I made was simple. AI can be impressive, but in business, impressive is not enough. AI has to be accurate.
This summer my daughter asked me if I'm still talking unicorns at work. She thinks that's what I do all day. So we drew another one together this time. A couple of things are different. First, the unicorn is standing outside of my office, in front of our headquarters. It looks actually amazing. I told my team we should make this a reality. Dominique. And maybe we have some budget for that. Second, it has actually two ears now.
We don't take any credit for this. Frontier models have gotten better. But here's the thing. Ask the model to draw a unicorn and you get a great picture. But if you ask the model which of your suppliers will deliver next week, it will guess. It still has no idea how your business runs. And that's why this time we didn't draw the unicorn in the chat window. We did it in Joule Work.
Because Joule Work does both. It can draw you a great unicorn and it knows your business inside out. Since Sapphire, we have been working with hundreds of customers to implement Joule Work as their new AI engagement layer. In fact, some of these customers are with us here today to share their outstanding feedback across every line of business.
Starting this month, we are rolling out Joule Work to all of our customers for SAP and you. This is a big moment and a new UI for AI. For decades people had to learn how SAP works. I'm sure many of you can remember. I certainly do. Memorizing all the transaction codes. If you are lucky, you can export data down into an Excel spreadsheet. You have to build PowerPoints. End users did all the work and told the system what to do. A very manual way of working.
But now with Joule Work, the more than 400 million SAP users can talk in their own language to SAP, and Joule Work gets the job done. Joule Work is based on three pillars. First pillar: conversations. Here you can ask questions, delegate tasks, and truly route it to the right agent. The second pillar is Spaces. Here Joule can visualize SAP and non-SAP content, analyze your business to make smarter decisions, and help you collaborate with your team.
And finally Joule Studio. Here you can extend SAP agents end-to-end or build your own agents tailored to the needs of your business. Joule Work — once on your desktop, on your phone, and inside the tools you're already using today. Behind the simple experience sits the most powerful business AI agent engine in the industry.
Joule accesses our secret sauce, the very deep business and data context of SAP. Our Knowledge Graph maps more than 7 million data fields, half a million tables, over 5,000 APIs, as well as our private APIs, 400 data products, and our SAP process knowledge. Other AI coworkers, yeah, they understand some data of your business — data that is sitting in a PowerPoint, in an Excel spreadsheet, or maybe in one of your mails in your large inbox — but Joule truly brings everything together to understand the full context of your business, including SAP and non-SAP data. And that's why Joule doesn't guess.
Joule Work is also live at SAP for our 110,000 colleagues. You can call them. They use Joule to perform hundreds of thousands of tasks every day. We are already seeing productivity gains in central functions like HR, finance and procurement of over 20%. We see even more productivity gains when you have to get done with more complex analytical or planning tasks.
Joule has the power to transform every role, including my world. The role of the CEO. Let me now give you an example from my life. Eric, water. If I like it or don't like it, be prepared for earnings. Depends on the numbers. In the past it used to take three weeks, some back and forth between finance, investor relations and communications to get everyone ready. Now with Joule, what used to take weeks can be done in days.
So what are we doing differently? I start by asking Joule, hey, can you help me to prepare for my Q2 earnings call? Joule now pulls together our internal financial data, some external analyst information, and provides me a really concise summary of the quarter. Now with Joule Spaces, actually I can see now a dashboard with all of my KPIs relevant for earnings, including some benchmarks to analysts.
And finally, Joule actually also recommends me anticipated questions, what analysts might ask, and gives me the key messages to the market. And finally, I always do that help me choose the right messages so that the share price is going in the right direction. Now Joule is good but not sure always share price goes in the right direction. Honestly, if Joule could now also run the earnings call, who needs a CEO in the future?
And there's one final important thing about Joule Work. Joule Work is open. What does this mean? You can call Joule via our Agent-to-Agent Gateway from any AI and get the same governed answers from SAP. And the same is true in reverse. You can discover and call any third-party agent registered in the SAP AI Agent Hub.
Now, for certain tasks, end users need to do more than asking a simple question. When you have to move money or run a payroll, a good answer is not enough. You need the right answer every single time. The answer must be compliant and traceable. This is where our AI assistants and agents from our Autonomous Suite come in.
At Sapphire, we reimagined our applications portfolio into five autonomous domains: finance, spend, human capital management, supply chain, and customer experience, as well as industry AI. When you need to execute a certain task where accuracy is a must, Joule Work selects the right AI assistant for the job. Think about an AI assistant as your best virtual buddy when you get stuff done in Joule Work — an AI assistant, a domain expert for sourcing, for HR operations.
The assistant orchestrates the specialized AI agents that do the actual work. But how can our AI assistant guarantee the right answer when LLMs are probabilistic and only guess the most likely answer? Now, because our AI agents are embedded in the deterministic workflows of our applications, they deliver the one right answer because they are running on the same data, they follow the same rules, and they leave the same audit trail as all our applications.
And you decide on the level of autonomy — human in the loop when you see a bigger risk, but you can also deploy the agents fully autonomous with guardrails baked into our applications. Since Sapphire, we have made more than 20 assistants and over 200 agents generally available, aiming for more than 400 agents by the end of this year.
But the real story isn't the number of agents. It's the business outcome they deliver for you, our customers. And our AI assistants can only understand your business so well because of the business AI platform. The business AI platform is the place to be. It's the place where you build, contextualize, and govern AI for your entire landscape.
The business AI platform consists of three pillars, starting with build. Here, Joule Studio gives you a seamless handover from no-code to pro-code, giving you the flexibility whether you want to extend an agent or build a new agent. We also provide access to over 70 models to avoid lock-in and optimize your token consumption.
Second, context. This is where the brain of our agents is sitting. Our Knowledge Graph and Business Data Cloud give our agents the business context and the ontology out of the box. Again, important: SAP, non-SAP, and third-party. Third, governance. Every agent you build on the business AI platform is one and governed by SAP. Very important: one and governance by SAP.
SAP ensures for every AI agent identity and authentication management, compliance with regulations in over 60 countries and 26 industries, and certification by your auditors. And with our AI Agent Hub, we provide one inventory for every SAP and non-SAP agent in your company.
So I know I might be a little bit biased, but I have to ask you one question. Why would anyone build agents stand-alone on an LLM platform? The business AI platform gives you the choice between the best frontier and the best open-weight models for every agentic AI use case, plus the deepest out-of-the-box context, plus SAP-managed governance. It gives you the best business AI for enterprises.
Now I have talked enough. Let's see it in action. Over to you.
Thank you. Christian. Let us look at the SAP Business AI Platform closely as Christian laid out. Joule Studio is natively integrated with our context layer. So the agents you build are more accurate, faster to deploy, and more efficient to run.
Our Knowledge Graph is the map with the processes, data flows and business rules. So the agents find the right processes among thousands and the right information among our 7 million fields. SAP Business Data Cloud brings star understanding to all your data. One consistent view of the processes, policies and logic that make it meaningful.
SAP HANA Cloud and Databricks give governed access to distributed data. And Reltio harmonizes master data into trusted profiles of customers, products, and suppliers. Our domain models trained on SAP's code understand your business rules for predictions. There are two engines: our own model family trained on SAP data and Priya Labs Tabular AI, both grounded on how your business runs.
Taken together, your data, processes, relationships, rules and master data live in one complete business context so people and AI share one understanding of how your business operates. From that, the potential to transform your business is limitless.
During a merger, bringing an acquired employee onto your system costs at least 20 hours of work time and dollars. Now imagine if this is a 500% acquisition. There will be substantially large work and millions in cost. Let's see how we can actually do this with Joule. With Joule, the Autonomous Suite, and Business AI Platform, these hours can come down to a fraction of the cost by more than a quarter, and business users are in control.
Imagine Blue Peak just closed its acquisition of North Star, and 25 new employees start a week from Monday. We'll follow Anna, the HR business partner, and Brett, a developer. Anna opens Joule Work and her HR assistant already prepared an onboarding plan and a Joule workspace to run it from. The assistant started the moment the merger was signed and Anna didn't even need to ask.
That's the Autonomous Suite. The assistant recognized what an acquisition closed means across HR, IT, facilities and procurement, and it laid it all out for Anna with the same authorizations and audit trails. She looks at the new hires. Now North Star runs a different HR system, but with Databricks can easily connect to the North Star system and bring the employee records into Blue Peak SuccessFactors without losing context.
Now that's SAP and non-SAP data in one context. So the assistant reasons over the whole company, not just the part with SAP. Now Anna approves the office space, the badges and the office access. But she recognizes laptops are at risk. Anna can see only two of the 25 laptops will arrive by Friday. That's not okay. Joule recommends a fix: an alternate supplier.
The HR assistant doesn't have the capability yet, so Joule recommends extending it. She turns the recommendation into a development request. This is what Christian was referring to. This is Knowledge Graph at work. The deadline, the supplier, the approval rules — all attached to Anna's request. A standalone LLM would need all of this explained every single time.
Anna ensures any rush order requires human approval, and sends the extension request directly to the development team, which Joule Work confirms. Now let's move to Brett, who is the professional developer. And Brett receives a notification in Joule Studio. He opens the request and sees the extensive business context and the intent.
After review, Brett generates the solution. This is intent-based development. Anna wrote the requirement in business language. Joule Studio turned it into a working extension. The extension inherits its agent's governance, its authorization, and its context. Brett reviews the night and workflow: after the primary supplier check, one path selects an approved alternate supplier and creates a requisition.
If one doesn't exist, the other sends the request to a sourcing assistant for bits. Nitin is natively integrated with Joule Studio, so the workflow is something Brett can see and change. Now Brett runs the evaluation scenarios, deploys the extension, runs the required checks. Joule Studio confirms the extension is published in Agent Hub.
Brett opens the extended agent to review its governance record. This is the governance layer. Every agent, SAP's and users', managed across the life cycle in one place, and that too is SAP-managed. Anna's notified as soon as the extension is deployed as well. The extension found an alternate supplier on the network, with the stock on hand and the delivery prediction that clears the bar.
Anna approves it and the buying assistant executes the rush order. The space now shows what comes next. The onboarding plans for each new employee, one click, and the onboarding plan are shared with all 25 employees. Instead of chasing four teams, HR ran this from one space. That's the true power of the autonomous enterprise.
With the business AI platform and SAP context, it gives you clearly differentiated experiences without compromising governance and consistent accuracy. No public LLM alone can do this for you, and that becomes really clear when you compare Joule against the frontier models. In every product line, Joule is not only more accurate, it also lowers the cost.
For everything you just saw is ready for you to explore. I'm sure you're imagining how it applies to your business, but I think what will speak more powerfully are the stories from our customers. For that, Jan, over to you.
Thank you. Thank you. Yeah. First of all, thank you very much. And good morning. Let's wake us. I love what Minaj just said about the importance of execution and outcomes. And that's exactly what the SAP Autonomous Suite is all about. You heard it from Christian: five domains — finance, spend, supply chain, HCM, customer experience and industry AI. And this is not a vision anymore. It is already running in customer systems today.
But you know, today I would like to do something a little bit different, rather than giving you the same sales pitch for the Autonomous Suite that you probably heard a hundred times by now. So I've worked for SAP now over 20 years, and now I'm not resigning on stage. Don't worry Christian, or maybe don't get too excited Christian, but I have a confession to make.
Deep down, I've always wanted to start my own business, and during the World Cup this year in the Americas, I got really inspired and I saw a huge opportunity. Instead of selling software, I want to sell jerseys. I live in New Jersey. Can you tell from my accent? And since I live in New Jersey, I got the perfect name for my business. I'm going to call it Jerseys from Jersey.
All right. But that's a small problem. I've spent my entire career helping other people run complex businesses. I've never built one myself. I've never procured a product or ran a production or managed a supply chain, paid vendors or collected payments myself. I'm definitely unqualified for that. But I want to build Jerseys from Jersey into a world-class business.
So how about you guys raise your hands if you have ever run your own business? Who has run your own business? Who has? Keep your hands up if it's more than one. See, I see a few hands, so I think I will find some help here in the audience. But for now, I've actually asked a few of my customer friends to help me out here, because they really use the SAP Autonomous Suite to run their businesses, and they know how best-in-class looks like.
They have experience of using AI to take every functional domain to the next level. And they can certainly help me get started. So where should we begin? Well, let's start with procurement. Who doesn't like procurement? I need to get the right materials in the best quality, for the best price from the right supplier.
But before we go there, I have a quick trivia question here for the audience. And the reward will be actually one of my jerseys. So what everyday item can become the fabric in a professional football jersey? Is it A) plastic bottles, B) coffee filter, or C) tissues? Are you also quick? See an overachiever here front row. It is plastic bottles. Very good.
Okay, so let's go back to procurement excellence. Who better to ask than Novartis? Because they operate in an industry where lives are literally at stake. And when it comes to their procurement discipline, there is no margin for error. So let me bring out my first customer friend, the head of procurement at Novartis, Christoph Berkey.
Good to see you. So, Christoph, first of all, thanks so much for being here today with us. And I said, nobody operates in one of the most highly regulated industries in the world. As a head of procurement, what's the biggest challenge that you and your team face?
Yes. Procurement touches hundreds of categories and a huge number of suppliers. The challenge is to make decisions quickly and with confidence. Creating RFPs, analyzing bids, preparing for negotiations are all very manual and time-consuming processes. At scale, that creates inconsistency and slow cycle times. And in a highly regulated environment, this can be very costly.
So context is super critical. You mentioned that for you guys, especially when it comes to controls and a sense of traceability, auditability. What you have done is you launched the pilot now with autonomous to deploy autonomous sourcing agents on top of the business AI platform. What's your vision there and what are you trying to achieve?
Yes, that's right. We have co-developed and recently started a pilot using SAP's Joule Sourcing Assistant. And the focus right now is on bidding, analysis and negotiation support, with the event builder agent coming in a few days. Our vision is that agents handle the analytical heavy lifting, things like creating bids, comparing offers, and identifying negotiation levers.
But even at this early stage, what we are already seeing is that agents bring consistency and speed to work that previously required a lot of manual effort and coordination. We expect 25% faster cycle times and 35% uplift in productivity for the fights. Wow. Those are some impressive numbers. That's amazing. It sounds like a huge potential for real, measurable outcomes.
Can you maybe elaborate a little bit on some of the outcomes that you're building towards, and how you will know that you were successful at the end?
For us, success looks like faster and better-informed sourcing decisions, less time from requests to award, and with stronger analytical foundations and compliance behind each decision. We want our sourcing professionals spending their time on strategy and supply relationships, not manually processing bid tables.
That's amazing. Now, maybe from a little Jersey company here that's just getting started with autonomous agents, what's your best advice on where to start?
Actually, yes, we are in the semi-autonomous stage, but getting ready for full autonomy next year. In preparation, I suggest get your data in order first. If the underlying data is unreliable, the agent output will be too. Pick a process where a wrong decision is recoverable. Build confidence before you expand into anything mission-critical. Involve the people who do the job. They know exactly where the friction is, and do not underestimate change management. The value is not just in deploying, but also in where people change how they work because of it.
That's amazing advice. Thanks so much, Christoph. I think I do have a jersey for you. Here it is. Thank you very much. Thank you. Thanks so much.
Okay, so now I know how to do procurement from an expert. But the next step is really production. You know, actually making my jerseys, and just in time because I'm already seeing a demand spike. Maybe it's still from the World Cup. The finals were actually in New Jersey. And this is exactly what the Autonomous Suite was meant for. Let's see how it works. So could you please come out? Christina Richert. Here she is.
Thanks for being with us here. So how can I meet the demand, produce my matchday jerseys and get them to fans everywhere?
Sure. Yeah. But first. Good morning everyone. I'm going to take your jacket from my production side, if that's okay. And I'm ready to show you. So first, my supply chain agents bring together real-time sales, inventory, production data, fan trends, and they show me exactly which activities require my attention.
And yes, Jan, the demand is growing beyond what you hoped for, in what I planned for. The agents show me exactly where demand is growing and where we're falling short on those preorders. And there's something else. Quality inspections are showing that the prints are failing, so the agents immediately put a request for more fabric.
But the reality is, now I have two open fronts. On the one hand, I need to keep up with the quality, and on the other side, I also need to keep up with the demand. But don't worry Jan, we will make it so. The agents, as I mentioned before, they showed me that they already put in a request for more fabric and they're waiting for my approval.
But before I approve, I really want to understand why are these prints failing? The agents follow the evidence and they found the root cause. A new supplier's fabric does not meet our standards. The agents show me all of the inspection results. They also show us which jerseys we can fix and send to rework, and which ones we definitely need to scrap.
So in this scenario, Jan, we went from a clear warning into an analysis and understanding of what happened and what is the impact. And now we need to act. And yes, before you say anything, I know we need to act fast. So the next most logical step is to check within our existing providers who can help us as fast as possible with that fabric.
I am presented with two options: two suppliers with quality, quantity, cost and delivery time estimations compared. And I think so, supplier number one meets our quality standards and can deliver in one day. So I will approve in this moment. Procurement is set in motion across all of our manufacturing sites.
And now with the new supply, the agents update the production plan and we're ready to start production. But we are also, in parallel, preparing for our transportation and shipment plan. So demand went up, who already was at risk, but we are super ready to deliver. This is our tournament supply chain and business AI working together with agents across different businesses.
In this case, demand planning, inventory, manufacturing, production, transportation, the people making the decisions, and the business moving forward together. So the jerseys in the hands of your fans, that's the promise I'm going to keep today. That's really impressive. Thank you so much. Thank you. Thank you. Thank you. Thank you.
Good. So we've now covered procurement and production. But one thing that Minaj showed as well is the importance of people and keeping people in the loop. You heard it from Christian as well. So we all agree here in this room that people remain critical. But how do I make sure I'm hiring the right people with the right skills that I need?
It's a lot more complicated than just upskilling for AI, and the gap between the workforce that you have built over the last years and decades, and the workforce that you need going forward, is one of the biggest challenges for every enterprise. And to close it today, I'm super excited to announce that SAP is officially acquiring Tech Wolf.
So tell us a little bit more about that. I'm also very pleased now to welcome the head of product for SAP SuccessFactors. Rebecca. Rebecca. Come out. Perfect. So first of all, thanks for being here as well. We're super excited about it. It's actually a pretty cool name, Tech Wolf. Great. So what is Tech Wolf and how does it help organizations create the workforce that they need for this moment?
Well, first of all, thank you so much for having me. I am very excited to be here. You cannot redesign work if you don't understand what people are actually doing, the skills that they're using and what they have the potential to do next. And for most organizations, that picture is very incomplete and very, very out of date. Tech Wolf is an AI-native work intelligence platform built to address exactly that problem.
Their context graph for work brings together first the task on a job, then the skills people have and apply each and every day, and finally signals from the external labor market. Tech Wolf then connects all of those signals to your organization's goals, so that leaders can see how work is changing and what that means for their people.
Now, I'll say on a personal note, this is a really big, complicated problem. And when I first met this team, I was struck by how deeply they had a passion for solving it and understood the customer circumstance. They aren't just thinking about how to improve what exists today. They're actually trying to reimagine a different future for work intelligence. And that is exactly the expertise that we are excited to bring to SAP.
This is going to be critical if we're going to help customers make workforce decisions at the pace that they expect. That's fantastic. And what's the biggest benefit for our customers?
Well, the biggest benefit is being able to act on the understanding that I just described. Today, a business can decide to change direction. In fact, many of you are probably doing that today, and it takes time to see which work needs to change, what capability it already has, and where people could grow and where they also need to hire. Close to my heart.
These decisions unfortunately often happen in very different tools. Now, if we think back to what Minaj just showed us, Anna was bringing a group of employees into a company that SAP was helping her to onboard. But let's imagine taking that one step further as those people join the business. You can immediately understand the skills and experience that they're bringing, where they will fit in alongside the existing team, and who could grow into new work, maybe even where genuine gaps still remain.
And from there, that same understanding can now guide each person's development, pointing them toward the right learning, the right mentors, the next opportunity inside a company. And of course, when people can see a real path for growth, they're far more likely to stay. And that means that the business can keep building on the talent that it already has.
Now, that same understanding can also apply every time a team needs to hire a different set of skills. Now, maybe before looking externally, leaders should be able to see the capability that they already have. So when they make a decision, the next steps shouldn't have to be assembled manually, but can be seen right in their system today.
So of course, following the acquisition close, we expect to bring Tech Wolf's intelligence directly into the core of SAP SuccessFactors, carrying it through workforce planning, learning, mobility, recruiting and of course, obviously the broader SAP ecosystem. So it's very exciting. Awesome. It is exciting. And let's give them a great onboarding and welcome them to the SAP family. Thanks so much for being here. Thank you. Thank you. Thank you.
Good. So we covered procurement. We covered production. We covered people. But one thing I haven't even touched is my supply chain. And it's the most important thing for all. For any business, I would say I need a real expert. So let's ask someone from a company that's been doing it for over 125 years, and that's Morgan Foods. So please welcome their chief financial officer, Steve Hankins. Steve. Come on.
So, Steve, thanks so much for being with us here this morning. Now the supply chain for my little Jersey business is a lot simpler than yours. You make products that millions of families expect on the shelves, grocery store shelves every single day. How do you manage such a complex supply chain?
Thanks for having me here today. So at Morgan Foods, we produce hundreds of private label products across multiple facilities: soup, beans and cans, aseptic product in boxes, and everyone has to arrive on time on spec. But our complexity is really not just the volume. We have to coordinate forecasting, ingredients, production, scheduling, warehousing, distribution, and really all at the same time.
And before SAP and S/4 and IBP, most of this was manual in silos in Excel. When 2020 hit and the demand for canned food just literally exploded, I'll be direct. If Morgan had not built the SAP foundation that we built, I don't know what would have happened with us with S/4HANA. We connected our business with IBP. We continue to make it more predictive.
And really, for the first time, those things in combination enabled us to see our whole supply chain and begin to plan ahead of it instead of trying to keep up with it. So we orchestrate across forecasting, planning, supplier collaboration and execution and try to provide exceptional, exceptional service to our customers. Our service level is usually 99% plus.
Our products are very seasonal. There's a soup season in the winter time, and so we have to plan ahead of these seasonal spikes. We share our factories so that we could optimize the cost of our products. We try to optimize our inventories and really work that equation together. And IBP is how we facilitate that.
But a big thing for us and a goal at the very beginning of the process was this one number that says it all. Our response time to large customer requests was 3 or 4 weeks before we could determine if we could satisfy that or not. Today, we can do that in about five minutes. That's the difference between telling a retailer, we'll get back to you in 3 or 4 weeks, or yes, we can do that. We can take care of that.
Because when you're feeding as many families as we are, because our products are in every retailer. Yeah. You just can't have blind spots. You've got to have the whole operation connected. You got to have that context of where you're at and that plan, so you can make these calls quickly, and you got to have your processes in control. Yeah. No thank you. Thank you for that.
And I mean, you've been talking about connectivity between systems, mentioned IBP, a long end-to-end business process, which is actually at the core of the Autonomous Suite. And that is the foundation for you right now. AI agents, they are showing up everywhere in supply chains. Where does it take us now? And what's your advice for someone that's starting from scratch?
Okay. Good question. So our biggest shift is always moving from reactive to predictive. And agents can connect demand signals, inventory levels, supplier status. It does all that faster than any team or any individual can do. But we're early in the journey. But we're putting these pieces in place. We're leveraging our technical foundation. We'll continue to do that. And that brings us some pretty important lessons.
First of all, technology only delivers as you've got the foundation there, the right data, master data, the right business processes, develop them, have them clean, have them documented. And with that technology and starting fresh can actually be an advantage because you can build from day one with those things in place.
I like to say that technology enables business processes, but you can't forget the people side of this. We have over 800 people in our organization that's making all of this work, and the technology makes them a lot better because people participate in business processes, whether it's at the operative level or the oversight level.
Now, also, Morgan is in a small town, 4,300 people. And to attract talent in our business, we need to have good processes and systems. That's the expectation of the people that are going to be working in the future. And we feel like with what we're accomplishing in Morgan, we're laying that out so that we can get the right kind of talent. So we're excited as we make this shift.
I hope all of you are excited as you look forward to do that. And remember that you have to have people on top of agents. That oversight is extremely important. They can't replace judgment and AI can't have judgment. People have judgment and it comes from experience. I couldn't agree more. Thanks so much for being here, Steve. Thank you very much. You're getting a jersey as well. Thank you so much.
Good. Jerseys from Jersey. It's really coming to life. But what about getting paid? The final trivia question for you guys. U.S. dollar bills are mostly made from what material? A) cloth, B) cotton, C) rubber. Shout it out, A, B or C. It's cotton, 75% in fact, and 25% linen. So apparently I'm in the fabric business on both ends.
But back to payments. Since nothing is more important for a business right now, paying suppliers can mean approving an invoice in one system, initiating a bank transfer somewhere else, and then reconciling it all afterwards. Three steps for something that should just happen, right? Think about your personal life and how easy it has become there.
So to make it just as easy in the enterprise, I'm excited to announce SAP Pay. You can clap. Yes, SAP Pay. It's a payment service that's embedded directly into SAP Cloud ERP. And when an invoice is approved, the SAP Pay agents, they execute the payment. They reconcile it automatically right inside the ERP, supporting fiat digital currencies as well as cross-border payments.
So basically, after onboarding customers, they can make payments in various payment methods like domestic and cross-border wires, checks or even stablecoins. And then the natively embedded nature of SAP Pay cuts the implementation time from months to days. And we see a potential of 25% TCO reduction on average, which is quite impressive and something you should definitely look into.
But what about collecting payments that are due to me, which is a key part of having an end-to-end finance operation? You can't build anything with cash that's sitting in an unpaid invoice. So the faster you collect, the healthier of course your business is.
So now imagine you aren't running a small jersey business like me, but instead a global giant like Nestlé. Or like the American, say, Nestlé. You are. You're managing receivables across hundreds of thousands of retail customers in nearly every country on Earth. And your accounts receivable team is making decisions about outstanding invoices every single day.
And that's someone I need some real advice from as I build my collection operations for my little company. So I'm going to call in my next customer friend, Nestlé's group head of IT and digital innovation, Luca de la Lettre. Luca. Come out.
Luca. Thanks for being with us here today. First of all, you have just completed a record-breaking S/4HANA RISE conversion, so big congratulations on that one. And so what was the goal and how does Joule and the accounts receivable assistant now fit into that?
Well first of all thank you so much for having me. Hi everyone. The transformation wasn't about technology for its own sake. It was about strengthening our core, operating as one connected enterprise with consistent processes, trusted data, and greater visibility across markets and functions. That's been mentioned, but when data is fragmented across several legacy systems, there is very little to work with for AI.
It's only when the underlying data is harmonized and trustworthy that you can deploy any AI that can make a difference at our scale. And yes, for an hour allowed across more than 140 countries. Thank you so much for the efforts. It creates exactly the foundations as Joule sits on top of it, as the layer that enables AI to reason across processes because the data underneath is connected and prepared for it.
So what makes the difference is that scale versus deploying just point-to-point solutions. In our case, AI is not passed into a fragmented data landscape, but is actually operating on a clean and unified core.
Now going back to the collection agent, imagine this picture: there is an agent that actually prioritizes, automatically ranks the work list based on risk, relationship history, invoice value. There is a value and the most important actions are always visible at the top. It also recommends the next best actions that get smarter, whether that's a call, an email, an escalation, or a hold, and if outreach is needed, it also drafts the communication.
We are also young at adding a touchless order management system, so where orders come in, they get created in the systems automatically and they get fulfilled, so it becomes completely touchless. And as you can picture and imagine, this represents a big saving for Nestlé, especially when it comes to working capital. And they sell sustaining.
Well that's super interesting, touches and also directly or indirectly into customer experience. I have that later. But I hear definitely a red thread here, right. It's about connected processes based on lean governed data. But let's zoom all the way out. When you have the power of modern end-to-end operations, like the kind that you're building, what changes for the business as a whole?
Look, in few words, it's really the power of business and digital transformation at scale. In this case, specifically within the finance context, we can move from periodic planning cycles to something that is more continuous. Forecast can change as conditions change. So when the next planning window opens, financial signals to face so the vision can be addressed before they become problems, bringing faster response to change.
And imagine this when planning sits within the same unified foundation with business context. Every edge you add on top of it gets smarter because it gets more of that business context. That is the compounding effect. And one final point when you operate at our scale, it's even more critical to have your data connected with real business context. And then the contract governance it all. Awesome.
Well, I can definitely see how context is a critical element here, something that is not accessible for generic LLMs and one of the key differentiators when we think about the SAP Autonomous Suite. So thanks so much, Luca, and congratulations on these fantastic results for Nestlé. Thank you so much. And of course, and of course you're getting a jersey as well. Oh of course. Nobody lives without the jersey.
Good. So we've covered now procurement, production, people, supply chain, finance and payments incoming and outgoing. You heard that. But there's one domain I haven't touched yet. And honestly, it's one that underpins everything. It's customer experience. Because think about it. When Messi signed for Inter Miami here in the US, 2 million jerseys were sold overnight. That's my dream.
And that was more than just a demand spike. It was a massive moment. Everybody who opened their phone found the jersey, clicked buy, and expected it to show up fast and perfect. And for a company like mine, e-commerce is everything. So every touchpoint the fan has, from browsing to buying to opening the package, has to be seamless.
And what makes the Autonomous Suite different is that the same data powering my supply chain also personalizes the customer experience. You heard a little bit from Luca about that already. So from the storefront to anticipating what fans want next, it makes the experience just as good as the jersey itself. And it all starts with this.
So every step in this journey has been leading up to this. And that's exactly what turns a one-time customer into a customer for life. But of course, how you deliver that experience looks different depending on your industry. A Jersey company is not a pharmaceutical company, and every industry has its own version of customer experience and its own flavor of every functional domain we've talked about here today.
There's tremendous opportunity when you have autonomy with the deep context of your industry. So who better to speak with than someone who sees across industries? Please welcome the chief accounting officer and business services leader for PwC. Notes.
Thanks for being here, Jill. Thanks for being with us. So PwC works across just about every industry. And what's the first thing that changes when you move from generic AI to one that understands how a specific industry operates?
Thanks, John. Well, you know, every industry has its own language, its own risk profile and its own definition of what done right actually looks like. So, for example, a healthcare company navigates compliance, patient safety and highly regulated workflows all at the same time. And a transportation company manages capacity, routes and time-sensitive delivery windows.
Those aren't just different use cases. They actually require fundamentally different AI logic. So what we've seen at PwC with our clients is that people actually trust AI to do more when it really speaks their language. That means understanding the context, including what matters, what the consequences are, and what happens next in that specific process.
Generic AI can execute tasks, but industry AI understands why those tasks actually exist and what they connect to. Okay, so again, it is the context that is critical. And industry-specific context of course is extremely valuable, right. So now let's talk about your industry for PwC which is professional services. How is AI having the biggest impact for you guys?
Well, you know, our clients trust us with their most sensitive decisions. And that trust really only holds if the systems underneath are connected. The context is accurate and governance is built in. In professional services, every single decision affects the very next one. And the temptation is to think about agents just one step at a time.
But when AI has full context across an industry, it doesn't just execute faster, it actually makes better connected decisions. So, for example, billing in professional services is historically one of the most error-prone processes. And that can have a huge impact on our client relationships. That's why we're working together to build SAP's billing assistant that will guide our teams through that process.
SAP's billing assistant will actually help catch problems at the source, rather than cleaning them up after the fact. And frankly, the entire engagement management process in professional services is complicated and very time-consuming. Along with expertise and judgment, time is one of the most valuable assets.
As you know, we are continuing to explore with you SAP's industry AI working towards autonomous project delivery. So that will mean less time spent on engagement administration and more time to deliver better client outcomes, stronger client relationships and trust to do even more.
Well, thanks so much for sharing that. And I'm now the smallest business in New Jersey that got one-on-one counsel from one of the world's largest professional services firms. So thanks so much for that. And I have a jersey for you. Don't worry. The bill's in the mail. Thank you. Thank you. Take care.
All right. So I want to thank all of my customers, my customer friends here for their great advice. And when I said my Jersey company was hypothetical, I was only half kidding, because following this keynote, I invite you to walk out to the back of the show floor and check out the SAP Experience Center, where you can not only learn about the autonomous enterprise, but also pick up your complimentary jerseys.
And while we had a lot of fun today, I hope so. I want to end on a little bit of a more serious note. What you've seen today, the domains of the SAP Autonomous Suite are more than a collection of functional components. It's one system working together like connective tissue. And you heard all of our customers say it: end-to-end connectivity, context and control.
And by that I mean everything has to be secure and governed because anyone can build an agent nowadays. What they can't replicate is 50 years of business context and governance that is baked in: compliance, traceability, auditability. That's the autonomous enterprise and that is SAP built on a foundation of trust.
And to hear more about that, I am super pleased now to turn it over to Marielle Airman. Maria, come on out. Thank you very much.
Good morning, Las Vegas. So what you just heard from Jan and our customers is actually the best proof point I could personally ask for. It's real enterprise outcomes running live today. So thank you all for that trust. It isn't given, it's earned. So AI brings extraordinary opportunities to all of us. But technology in the wrong hands amplifies risk at a scale that we have never seen before.
Just a year ago, the boardrooms asked, how fast can we deploy AI? Today the question has shifted: how fast can we deploy AI without making tomorrow's headlines? So earning trust in the AI era means solving three things simultaneously. It's managing the legal risk. It's enforcing the enterprise security, and it's operating ethically.
And here's the reality we face. Attackers used to take eight hours to execute a ransomware handover. Today, automated AI does it in under 22 seconds. And while 83% of enterprises use AI daily, only 13% have true visibility on what it's actually doing. So the biggest threat today isn't the malicious AI. It's unmanaged AI.
At SAP, security wasn't added to our AI. It was never optional. SAP runs the core of the world's most critical business processes: its finance, HR, supply chain, procurement. Decades of handling mission-critical data and that forms the exact foundation this platform is built on for AI agents.
We apply zero trust, verifying the identity, the business purpose and data sensitivity on every single request. So GDPR, the EU AI Act, DORA needs to. These are all built into how we develop and how we are delivering AI. Compliance is part of the architecture from day one.
When people look at AI, they see what is above the waterline. The agent, the demo, the answer. But what determines whether AI can actually run in production sits below the waterline. It's regulation and localization, tax law, e-invoicing, labor mandates, data residency. Our platform understands these out of the box and applies them consistently across the regions.
It's identity and access management. Every agent operates under the exact authorizations of the person it represents. It can only access data and execute actions it's explicitly permitted to perform. It's data protection. Your data remains your data. It is never used to train foundational models.
And finally, it's governance and visibility. As you deploy hundreds of agents across your SAP and non-SAP environments, you must know who owns them and what they access and the risk they carry. That is the hard part. And that is exactly what SAP manages for you.
So let me be specific about what AI governance actually means. It is not a passive compliance checklist. AI governance is the live operational framework that answers three questions in real time. Who owns this AI decision? Who can see what is this doing right now and who can stop it? Without governance, you have speed without accountability. Governance is what converts AI from a liability into a competitive advantage.
So at SAP, every agent receives a verified owner, a defined risk tier, and scoped access rights before it ever touches production. We are independently certified to the ISO 42001, audited regularly, and follow a SOC-compatible development process built directly with our external audit partners.
So when your platform and your auditors share the exact same governance framework, compliance stops being a negotiation and it becomes a proof point. Our AI assistants operate within our and within your approval workflows. They respect your authorizations and they stay within defined boundaries by design, not configuration. Every action is documented. Every decision is traceable.
The autonomous enterprise does not mean humans are out of the loop. Our AI ethics policy mandates three levels of human oversight calibrated strictly to risk. That's the human in the loop for approval required before the action. We have the human on the loop actively monitoring and do real-time interventions. And then there is human in command for the ultimate override authority on high-risk decisions.
So the more autonomously an agent operates, the tighter the oversight is applied. And that is not a settings choice. That is an architectural principle. We call it the freedom within boundaries. So you are no longer asking should we use AI? You are asking, can we trust it enough to scale it? And a price, right? The enterprise is winning right now. Aren't deploying AI recklessly fast? They are deploying it responsibly and at scale.
And that is actually my wish. Let's innovate boldly together and stay secure together. Thanks so much. And back to you, Christian.
Thanks, Marielle. Trust is key to everything what SAP does and it has been for over 50 years. We are delivering the most trusted AI in the industry. And today you saw life here on stage. Joule Work. The Autonomous Suite. Our business AI platform.
But AI also requires significant change management to ensure it delivers the desired outcomes for your business. I know from my own personal experience driving transformation inside SAP that change management is one of the hardest parts of our jobs, and that's where our business transformation solutions are coming in.
Thinking of you twice that change at a business process level. This is where all truffles should start. Actually it checks, simulates and analyzes these business processes before and after to make sure AI is delivering the value you expect while providing the value quantification so that you can justify your total spend.
Linux automatically discovers and manages SAP and non-SAP agents running across your business. And WalkMe ensures people are using the agents in the right way at the right time in any workflow across any SAP and non-SAP workflows.
Finally, our AI migration toolchain supports every step of your cloud ERP migration with AI. From landscape optimization and data migration to deployment and ongoing optimization. The death of multimillion-dollar, multi-year migration spend. Oh.
And finally we have a gift for you. Gift. If end users don't want so much change on their UI because they are so much in love with SAP GUI, you can also use a more retro version of Joule Work with functional and yes, with a statue spa. No mouse required. Perfect. If your AI has to look like it's from the 90s.
I told Philip I'm still a little bit undecided which one I'm going to use, but I'm used to this actually. So look, what we have seen today makes one thing very clear. Joule is redefining how business is run. Of course, Nestlé, Novartis, PwC, Morgan Foods and many more.
There's a massive shift happening now from manual steps to connected workflows, from reactive decision-making to predictive insights, from humans in the spreadsheet to humans in and on the loop, all powered by our AI platform, ensuring that AI is accurate, governed, and extensible.
And most importantly, it's ready for you today so you can ask anything, sort of anything, build anything. This is no longer the start of something better. This is better. This is the autonomous enterprise. Thank you and have a fantastic SAP Connect.
