Andreas Welsch: 0:30

Thank you to SAP for sponsoring this episode. Today we'll talk about how leading enterprises are adopting AI to drive value, and who better to talk about it than someone who's actively working on that Bob Sakalas. Hey, Bob, thank you so much for joining.

Bob Sakalas: 0:44

Hey Andreas, how are you doing?

Andreas Welsch: 0:46

Doing well. It's so great to have you on the show, but I know not everybody might know you and what you do. Why don't you tell us a little bit about yourself?

Bob Sakalas: 0:55

Oh yeah, sure. I'm from North America. I'm part of the solution advisory, which is part of our customer advisory group. But we basically help customers get past the hype. And I know it's part of your opening here, right? But how do you take this great technology that's changing so quickly? How do you actually start applying it to your business processes? How do you become more efficient, more effective? And of course, we're all talking about AI because AI is absolutely the next big thing.

Andreas Welsch: 1:25

That's wonderful. That's exciting. I know you work at SAP and there's been a lot of conversation, a lot of buzz on AI as well, and you've been in, in the industry for quite some time. What's getting you excited these days?

Bob Sakalas: 1:41

It's becoming real. I guess that's where I'm gonna start, right? If you think about it, two years ago, every presentation I saw on AI was talking about what could go wrong. Oh my goodness. This Canadian airline sold tickets for a dollar, or somebody bought a Chevy Tahoe for a dollar because they tricked the chatbot into doing that. So two years ago we were talking about how this thing goes off the rails. Today we've got customers, like for example, we've got these BTP summit events. Actually they're now called the platform and Data Summits. We renamed it of course but, five different cities where we're gonna have customers presenting how they're using AI and other advanced technology to actually deliver outcomes and results for their business. So we've gone from what could go wrong and everybody was worried about it. To what can go right and how we're actually able to use this technology. And it's not SAP doing the talking, it's customers showing up and saying, this is what we're doing with the technology. This is delivering results for us.

Andreas Welsch: 2:43

That's awesome. And I think especially this part about practitioners and leaders talking about how they are using it in their business is so incredibly important because there is so much talk, there is so much noise in the market. And for me personally as well this year, events that I've attended where you heard from, real people, how have they approached it? What worked really well? What were some of the struggles they went through and what are they achieving at the end of the day? I think that's the most meaningful and the most valuable time that, that you can spend. Learning.

Bob Sakalas: 3:13

Yeah. And the demand's incredible. By the way. We went from three years ago we had two cities. Last year we had three cities. This year we have five cities. So we're gonna be in San Francisco, New York, Chicago, Toronto, and Atlanta, all in September and October. It's gonna be a great event. I hope you can make it.

Andreas Welsch: 3:33

Yeah, I hope I get to be there for, one of the events and, definitely hearing what your customers are doing and sharing. But, we've, heard so much in, in recent years ever since we've talked about big data, you not only can gather the data, but you almost need it, or you already and, always needed if you want to do AI and if you want to do good AI. And to me, that was one of the big realizations that, I saw many leaders go through early on. ChatGPT, LLMs everybody said, yes. Great. Now we can do all these things we haven't been able to do before and write new language and summarize information and write code and whatnot. But very quickly, people said, it's actually pretty bland. It's pretty generic. It doesn't know anything about my business. It doesn't know anything about my products, it doesn't know anything about my services and what I do. And we quickly came back to the realization that, hey, guess what? We need data, we need business data, we need good data and quality data and things that are specific to combined data. Yeah. Yeah. And so we've, been talking about this for, years, and I know you, you've been working with some of the largest brands on the planet. How are you seeing companies actually use this and what makes a difference for them in their adoption?

Bob Sakalas: 4:47

It's really a big wake up call. If you take a, and I don't care which LLM we're talking about, if you take any of the LLMs and you ask it a question about SAP tables, for example, what we discovered very quickly is that sure, the l LMS great at the conversation, but when you ask it a very specific question, you're gonna find out that it's maybe 70% right. And 70% right might be good enough for hand grenades, but it's not good enough when you're trying to figure out when something's gonna deliver to your customer. And so finding a way to take a knowledge graph and make sure that you absolutely have your current data combined with understanding the business process under, with having that conversational I, I don't wanna discount, by the way, I don't wanna discount the importance of the LLM. But we always had a usability problem, right? We didn't, we knew AI had great potential, but we had that little department in the corner that worked on data science and they worked on one problem at a time. Suddenly, it's now available to everybody. So there's a great big bang moment that large, language models have done. The usability factors incredible, but we also need it to be accurate. The thing that I think we all have seen, and by the way, it's not just when I say customers have now realized. They have to get the data right, if they're gonna have great possibilities with AI full stop, bottom line. And so what's great for us as SAP is that now the customer is realistic. The customer realizes, oh, I can't get the great outcomes where I'm getting 95% plus accurate answers on everything without having the underlying data be right as well. And incredibly I hate to say it this way, but the planets have aligned. We've come out with some technology recently it's called Business Data Cloud, but Business Data Cloud helps us get that data right and break down those silos so suddenly you can layer AI on top of great data and accomplish great outcomes.

Andreas Welsch: 6:48

Great to have all of that is, is coming together and how are you seeing your customers first of all, make the decision what to move forward with and what are the. What are they building on top of it? It's, one thing to, to have lots of business data in your system or maybe to, to some extent locked away now that you can unlock it and that you can use it. What are your customers using it for?

Bob Sakalas: 7:17

Everything. Let, lemme start with a very simple example just because I think it, it communicates the power of using AI in such a way that it not only. Helps your own employees, which is something we all wanna do, but it also helps the customer get a better experience as well. Syntax is a company out of Montreal. They're a consultancy that works on a lot of ERP projects. And so they had a it's gonna sound mundane, but I think it's an incredibly important story. They wanted to have their consultants use AI on their phone. They basically report their time. Let's say I just spent four hours working on my Andreas project for the Andreas company. Okay. And the AI would say, okay, I know that you have worked on this pro, these two projects in the past, and might ask me one additional question that says, are you working on project A or Project B? But I'm gonna now book four hours of time against the Andreas project. And so it's suddenly it turns out consultants might use pieces of paper or they may use all kinds of different things to jot their notes on what they're working on. They wind up putting it in later into the system. And of course there's, inconsistencies, there's billing problems, all that kind of stuff. So now the consultant isn't happy'cause, he or she's gotta go back and do it later. And the customer may not be happy if they think their billing's not straight. And so in this case, an AI interface was created by using SAP App House, by the way, is this design group that helps you design cool new things. I've probably been talking to the users, and so they designed the interface to connect to S/4 Public Cloud through APIs. So S/4 Public Cloud stays pristine exactly the way that SAP shifted into the cloud, right? It can continue to upgrade, but because it's using an API between the AI and S/4 Public Cloud, suddenly it just works and it works well. So a great example of something that you might think is small, but when you multiply that across thousands of consultants working on thousands of projects and everybody's happy, the customer's happy, they get great billing or correct billing, I should say, the consultant's happy because this become as easy as talking to your phone. A, great outcome and just one example of how AI is really making a difference.

Andreas Welsch: 9:44

That's really powerful. And, I remember for years we've also talked about it's these parts of a business process these tasks that people like in this case, consultants go through on a daily basis that are mundane before you innovate and before you change and they can be accelerated and significantly improved afterwards. So to me that's a great example of how AI connected with SAP can drive a real impact.

Bob Sakalas: 10:14

And one of the things we're talking about is, of course SAP is accelerating our innovation. We're doing that all in the cloud. So things like Joule and embedded applications are all gonna be part of our newest S/4 stuff. But we still have a lot of customers that are still on ECC that are looking to make that journey to the cloud soon, but they don't wanna wait for another two years while they make that journey. I'll give you another story, but it's a, it is a powerful one. I think Louisiana Pacific, I dunno if you're familiar with them, but they have building products, lumber, all that kind of stuff, big bulky stuff that they ship to build your house, basically. They, their system was still on ECC, but they wanted to take advantage of an optimizer to optimize their shipping. It turns out that when you've got. A dozen manufacturing plants and lots of their warehouses are actually yards with huge lumber and things like that. Optimizing the shipping the, fulfill an order that may have a thousand line items is really important to them because that's where all the profitability is. So low margin business, it's really big, bulky stuff and they has to get there on time or construction gets delayed. So they used, in this case, our business technology platform to do the plumbing. They said, okay, we're gonna take these orders multiple times a day out of the older ERP, the ECC system, use BTP to do the plumbing, to go to this third party optimizer that kind of did three, three dimensional chess. It wasn't an if then, if, then, if, then, like we used to do back in the nineties. It looked at all the orders and said, what's the most efficient way to build these orders, package these orders, shift these orders, and get'em to the right place at the right time so that the customer is happy and we've optimized the cost of that in our low margin business. That matters a lot. Riner, who's the, main leader there that put this project together? He said, we got it done in six weeks. Wow. And it contributed 2% to our profitability. An incredible success. And what's interesting about this project is as they move toward S/4, all this infrastructure that was built in a loosely coupled fashion with business technology platform, we'll be able to plug into their S/4 system then too. So there, it's not, it's no regrets work. It's work that they've done that contributes to profitability today, and it'll work in the future when they're now in the cloud with S/4.

Andreas Welsch: 12:36

That sounds like another great example and especially the part about being able to continue using what they've built and just connect it to their new SAP S/4HANA system. I think it's really there.

Bob Sakalas: 12:48

There's no reason to wait. SAP would love for you to upgrade the S/4 in the cloud right now because you'll get the best possible innovation and there's no doubt that you will. All our Juul which is our chat bot, that's the smart chat bot that knows not only the information around the process, but how to run transactions and all that stuff that all runs within. S/4 on top of Joule. We also have embedded applications that we're creating to make sure that processes run better. But BTP and gener generative AI hub will help you apply AI today if you're not there yet. So you don't have to, you know how many cus how many companies wait two years to start doing innovation that their users feel? That's not the right way to go. The right way to go is to loosely couple that and start innovating today.

Andreas Welsch: 13:34

Awesome. The, part about the urgency I, think is important because things are moving so quickly. There's hardly a week that goes by where you're not seeing any kind of big AI announcements or new models, new capabilities, new companies on one hand being formed, the ones that might be competing with you that you don't even have on the radar yet, or the ones that are incorporating it and that are becoming so much more efficient and effective into that point about. Connecting data that you have in your business systems with something that you might build as an extension or as an addition to to that core system and, data. I think there's a lot of opportunity, and I know for the last 10 years companies and the largest companies on the planet have been looking at things like, how can we get better insights from our customers? I remember working with retail customers and

Bob Sakalas: 14:29

I think you're bringing up a very important point, which is. In the old days, all roads led to Rome and today all roads lead to the data. We have to make our decisions faster, but when I ask customers are, you, do your decision makers get their data in real time today? The answer is no. Yeah I then ask them, do they want it in real time? And it it's unanimous. Everybody's absolutely. We're being asked that all the time. Guess what? We start talking about the future and AI agents and agents working together. An AI agent's not gonna be effective unless it's got a real time or very close to real time view of your business. If you're working on a batch process that ran last week, that AI agent can't make those recommendations or decisions for you in time for you to impact your earnings. And so when you start thinking about the, what's I've been asked this question a bunch of times, but what's different about SAP's approach to AI and most other companies? And so it's easy to compare, like one of our closest partners, by the way, is Microsoft, but Microsoft and Copilot, which I have on my own desk. It's really great at personal productivity stuff. I can write a long email, I can throw it in a copilot. I can say, Hey, can you make this email harder? Hitting professional and short. Instantaneously I get the answer it's a personal productivity thing. SAP is not focused on as much that personal productivity in its own little microcosm of me writing emails or me creating a PowerPoint. What is, what we're focused on is this idea of how do we make your business process really hum? How do we make your business process ship the goods? Half a day faster. How do we make that business process make recruiting people and landing them so we get the best talent better, right? So we're always focused on the business process, and we think that's the area that's gonna contribute the earnings the most. So if you want to deliver for your shareholders and show that your company's becoming more profitable and using AI in a very substantive way, I think the a, the AI focus that SAP has, which is on your business process. Every company, I don't care. I used to work on Walmart as an example, and even at Walmart when I actually made the sentence or I said the sentence out loud in the meeting, I said no matter how large you are, you really are the sum of your business processes. Everybody in the room would not at the end of the day, it's not the one-off thing that you did that makes a difference. It's the thing that you do every single day, and you want to do that really well. And I don't care if you're Walmart or if you're Bank of America, or if you're Home Depot or any of these companies. Doing the every single day process is what's gonna make you more profitable, is what's gonna drive your stock Price is what's gonna drive growth, it's what's gonna drive opportunities for employees. It's not that I wrote an email twice as well, and now I can go stand at the coffee machine and chat about last weekend a little bit longer.

Andreas Welsch: 17:53

And I think that's the important part, right? How do you use it to drive measurable business results and business value and, being able to do this in new ways now. Coming back to the retail example, I know customers and companies have been using big data and big data platforms for a long time. They put analytics on top. You could analyze your SKUs and demand on a store level on a city regional. State level and what have you, how are you seeing that changing now? In, in terms of requirements, but also in, in terms of what the, underlying capabilities look like when you bring in data from different systems, when you layer AI, generative AI, agentic AI and these things on top. How do things change and how does that drive the, value discussion in, the also specific

Bob Sakalas: 18:45

Yeah, specifically to retail. It's always been a struggle over the volume of the data, right? So you, let's say you're a category manager and you're trying to improve the milk category in your store. You would, try to get the right product mix to drive to make milk as a category, the most profitable that it could be in your store. There was a mathematical disaster of a problem to do great category management. And why? Because it took a lot of human power to analyze that, to fiddle with the different parameters, to see, oh, I need to stock more quarts and I need to stock more, half and half. And all of a sudden people are down. Now everybody's lactose intolerant. When did that happen? But but the whole point is that it was a very difficult thing for the category manager to make sure that he had the right mix to maximize the profit of that. And of course, this is happening with detergents, is happening with cereals, is happening with everything else. So now if you think about the AI age, we've always been drowning in too much data. We've always not had enough time to analyze it, and the people doing the analysis often spent 80% of their time just gathering and cleaning and making sure that their data was clean enough to do the analysis, which they only did 20% of the time. What we're trying to do with this, I think the term is data product economy, but the, basic idea is you define. Great data products that have all the metadata around it. So now you're doing less of your time stuck trying to massage the data, and you're spending a lot more time analyzing it. That's the first step. Now, AI agents can often do part of our job. Not all of it. Let's face facts. AI is absolutely not curious as a technology. There's no curiosity, right? It's not very good at white space. Because it's very good at doing tasks that it's seen before, but when you give it, what's missing in this equation? AI struggles, right? The curiosity is not there. The understanding of white space is not there, but now it doesn't tire at a lot of these topics that we just talked about. So as a category manager, what that, let's just pretend I'm at Walmart, that milk might be very different in Denver than it is in Phoenix, than it is in Canada, as an example. But I would get tired of the thousands and thousands of stores and the thousands of situations that I had that might be a little bit different. The AI agent never gets tired. So when it comes to that, I'm gonna gather the data for this specific market area. I'm competing against these other grocers that are using milk maybe as a loss leader, right? So there's unique competition by market space as well. The AI agents don't get tired. So all of a sudden, the things that we can do in a repetitive fashion, I become a much better category manager because I can now get recommendations that are very specific, the specific markets, the specific competitive conditions, the specific regional differences, I guess would be a great way to put it. I'm able to do a much more granular job because that AI agent's not getting tired. The data is cleaner, so I'm spending less time there. The agent's not getting tired, and guess what? I actually get to spend time on the white space. I get to spend time. I don't. Wow, that's interesting. One of the things that SAP's doing right now that's really cool is that we've got these smart insights that are part of our analytical tools where it'll go find mathematical or data correlations, I guess the right way to put it, that are different that I may not never thought of. I don't have to go ask the question, do you see something funny in the data? It can, it'll come back to me and say, Hey, we spot an interesting trend in the data. So then, wow. I, I'm excited because I just got a new insight that I didn't necessarily generate. So AI is absolutely gonna make a difference at the granularity level. Things like business Data Cloud are gonna get us out of this data, moving smushing, transforming, eing, whatever you wanna call it, and just deliver us good quality data so we can spend more time on the stuff that really hits a home run.

Andreas Welsch: 22:56

Beautiful. That's, the kind of thing we, we should all be striving for in, in our businesses to, to look for these kinds of opportunities. Yeah. Now you already mentioned a, couple examples with Louisiana Pacific and, others. Who else are you seeing doing this really well? Bringing AI into the business, driving more value from AI in their business processes and with their business systems.

Bob Sakalas: 23:20

I don't have a retailer that's gone public yet, because I think they're all looking at this as a competitive advantage moment. And they're all working really hard to use AI in these kind of areas I just talked about. But I do have a CPG example that's really interesting. CPGs consumer products in case not everybody's a CPG person on this call, but constellation brands. If you're familiar with them, they've got Modelo as a beer, they've got Corona as a beer. They've got other spirits and vodkas and things like that, but they're one of the fastest growing CPG companies in the world, and they have been for the last 10 years. And I don't know if you saw, but Modelo actually, I think's now the number one beer in North America, all that beer gets shipped from the south to the north on trains. Turns out that they're shipping 400 million cases of beer. Per year via train to North America. So I guess we're drinking a lot of beer. That's a lot of beer. That's a lot of, it's a lot of beer. When you consider, we got 350 million people and you're shipping four oh million cases, and that's only one company. We're drinking a lot of beer in North America, but that's a different topic anyway, turns out when you're shipping so much beer, there's a lot of damage on these trainings to the tune, believe it or not, of about a thousand insurance damage claims per week. So Constellation would have to the, way this used to work is it would take the person that did the receiving would take their phone. They would have to take 15 or 20 pictures of the damage, they would then have to create his damage claim. They would have to send it in. It was like a 20 to 30 minute process. Every single time it was damaged. And this happened a thousand times a week, 52 weeks a year. They worked with us and said, look actually this started out, they weren't even thinking AI, but it led to AI, right? How can we make this process better? And it turns out that these receiving areas where they unload the beer already had a lot of cameras. And so they could gather pictures, attach'em to the AI created insurance claim. They would have a person look it over to make sure that it looked right. They might have to add a little bit to it, but they went down from 20 to 30 minutes to every claim is now below five minutes, and so that process, if you think about it a thousand times, has 52,000 times a year. They now have a much better insurance claims process that's worth millions of dollars. This is not small change. So another example of SAP working with the customer to ensure that the customer has a better running business process. I can't stress enough. That's the SAP difference when the business process. I love that. Yeah. When the business process runs great, it drops to your bottom line. It's not just a matter of freeing people up so they can jump in the traffic 15 minutes early to go home.

Andreas Welsch: 26:22

And a lot of times I feel these initiatives might start as we need to figure out something that we can do with AI. But I think once you've really start with the business strategy, the business process and, you look at where opportunities to, to drive more value to, to automate things to. Also measure what the outcome is before we even start. Be clear about what is our KPI that we want to drive? What are we looking to improve? And you attach that AI project and the value to it. All of a sudden it's measurable and it's tangible and it's aligned with your business goals. And it's much easier conversation than saying, let's boil the ocean and look for some low hanging fruit and quick ones. Whatever buzzword.

Bob Sakalas: 27:02

The best way to start, no doubt, is a lot of these started with a workshop. A simple workshop that had the supply chain people and the IT people that worked on supply chain together with SAP to just have a conversation, to just have a design thinking session that what are the top three things we could, really make better around here? It turns out that AI is almost always an answer that can help. And unlike the old days, by the way, we, oh, we're gonna create a custom application. It's gonna take us two years to get it done. Because the AI is so multipurpose now. It's the AI helps us get to the finish line quickly. SAP's jobs to make sure that it stays relevant, reliable, and responsible. Because if it gets you to the finish line quickly, but it's only right 80% of the time, that's not a good outcome. And so starting with the workshop's a great way these events that we talked about at the beginning, by the way. These are gonna be customers talking about innovative problems that they are solving in these five cities. What a great way to start your thinking. You can go to a session about how did you guys remarkably change your financial planning? Same basic concept, right? Getting inspired by what another customer did is a great way to start.

Andreas Welsch: 28:18

Absolutely. And like I said, hearing from others how they have approached it and why they even came up with the idea and what it drives in terms of meaningful value. I think that's one of the big, benefits. I'm definitely looking forward to attending some of these events. You said there were five in, in North America in September, in, in October hitting all major cities and hubs. We'll make sure to put it in, in the description as well. So please do take a look at...

Bob Sakalas: 28:47

they missed Dallas. They missed Dallas where I live, so I'm upset about that. So I would argue that they missed one location. But yes, we got five great locations this year.

Andreas Welsch: 28:57

You said they were so popular that SAP has been expanding them over the last couple years there's always a chance that 2026 Dallas will be on the list.

Bob Sakalas: 29:06

We can only hope, but but yeah I would very much invite, and by the way, this is a free event if you're a customer. A lot of events, people are charging you entrance fees and everything else. This is a free event if you're an SAP customer. Worth attending, there's no doubt. But it's a great way to start your AI journey to think about it, right? I would also encourage people to reach out to their SAP account team and just say, Hey, I'd like to have a workshop with my finance team, or I'd like to have a workshop with my supply chain team, or my HR team. We, have been doing these half day workshops that have really been. I hate to let make it sound this way, but they dropped to the bottom line this year. How many times in business do we work on something that doesn't? Get us the attaboy this year, but you can be a hero at your company this year with AI

Andreas Welsch: 29:57

I think that's a great point to close on. You can be the hero of your company with AI this year if you work with SAP and you look for these meaningful examples that derive value. Bob, it's been a pleasure having you on and, hearing from you how you and how SAP are addressing these big challenges, how companies can drive innovation and try drive real value with AI today. So thank you so much for joining.

Bob Sakalas: 30:20

Great seeing you, Andreas.