We're just delighted to have SAP participating in the Citizens JMP Tech Conference this year. To my left is Irfan Khan. He graduated from the University of Westminster in 1991 in Computer Science. Was at Sybase for a long time, right? How long were you at Sybase?
18 years.
18 years at Sybase. Then Sybase was, I think, I wrote a note on this at some point. Sybase was acquired by SAP in 2010?
That's correct.
Yeah. And so what happened after that? So when it was bought, yeah.
So yeah. So I'm actually one of those fortunate people that have actually only ever given one job interview my entire life. I left university, joined Sybase, spent 18 years there, and then the company I was working for got acquired. So literally one job interview. 30th anniversary for me was last year. I did take a sabbatical. Thanks to my team, I was able to take three months out. So yeah.
That was your 30th anniversary.
My 30th anniversary.
That's amazing. That's amazing. So when SAP bought Sybase in 2010, what was going on in the database world?
Yeah. I mean, the database world was becoming, dare I say it, commoditized. I mean, there was a lot going on. In fact, one of my quotes that I tend to use is, "If you look up Carnegie Mellon University," I mean, they're, of course, a decent institution in its own right, but they also manage a database of all database vendors. And that vendor list is like in the thousands now. And I would say that that list, even in 2010, was still a pretty long list. But it's fair to say that the majority of vendors out there were still struggling with just performance and TCO. That's really what it was down. Feature functionality was pretty much complete, at least given the scope of most applications.
Oh, and Irfan, I completely forgot to read the Safe Harbor, which I'm going to do now. "During this presentation, SAP will make or we will make forward-looking statements which are predictions, projections, and other statements about future events. These statements are based on current expectations, forecasts, and assumptions that are subject to risks and uncertainties that could cause actual results and outcomes to materially differ. Additional information regarding these risks and uncertainties may be found in SAP's filings with the Securities and Exchange Commission, including but not limited to the risk factors in the risk factor section of SAP's 2023 annual report on Form 20-F." My apologies. Anyways, okay. And then in 2021, you became the President of the, you took this current role, right?
That's correct.
So, explain through us what was going on. What's the transition?
Yeah. Look, I mean, it's a very simple story. I joined, as I said, 18 years in Sybase. I spent a large amount of my time with computer science background culminating into a career that resulted in me becoming the CTO of the company, the first and last because SAP was then acquired. Sybase was then acquired. Having spent a significant amount of my time in development, I really was not anticipating staying at SAP for that much longer. Not having clearly understanding of the app space, more in the technology and infrastructure space, I assumed it would be maybe a year. I was positioned essentially internally, right, as a domain leader in technology. And I said, "Why not take advantage of that?" So I actually completely transitioned into a new career, becoming a chief revenue officer, so for the whole of technology at SAP.
So that was a kind of a big bold move on my part, but then equally so, a second career chance. So having spent the last seven, eight years in sales, go-to-market, when opportunity came about again to go back into development, I took the opportunity back again.
Yeah. And then how does the development of the HANA database fit into all this, for people who don't know?
Yeah. So HANA is one of those key, I guess, inflection points for the company. Back in 2010, approximately the time when I joined SAP, SAP had taken a journey which is predominantly AnyDB customer, select your own infrastructure, whatever your database, core databases may be, Oracle, Microsoft, IBM. We will use those databases underneath the applications as a system of record. That in itself was great for a period of time, but then we realized that there's a whole level of inefficiencies that goes with that. So for instance, if you wanted to run analytics and you want to run advanced, complex analytics, that's not necessarily conducive to running against live transactional systems because then you've got a workload profile which is more transaction-oriented.
If you immediately start doing large levels of analytics on top of that, then you're no doubt going to end up with some kind of a compromise that you're going to have to make there. The foundation of HANA was really conceived to be able to have those mixed workloads, the transactional workloads and the analytical workloads. It was all based around the new thing at the time, which was in-memory computing, and having a huge level of efficiencies driven because it was also driven by compressing a large amount of data. Advocacy was take your very large, significant ERPs from an on-prem perspective, recast it into a HANA foundation, which gave you efficiencies of storage, but also huge levels of computational benefits from that as well.
Today, if you are moving to the cloud with SAP, can you only run on HANA?
Yeah. So SAP S/4HANA, which is the application suite of SAP, is natively built on the HANA technology. So it's not a case of do you select HANA or not. I mean, firstly, it's a SaaS application. And more often than not, customers care about the various SLAs, KPIs that they want to run within the SaaS stack. It's less about the underlying technology or the infrastructure and more about the value add on top. So answering your question, SAP S/4HANA is built around HANA technology.
Yeah. Okay. So two questions, I guess. First of all, and I usually start with this one, but so and you didn't report that long ago, but how's business? How are things going right now?
Yeah. I mean, look, SAP revised its guidance for 2025. The business is doing exceedingly well. The benefits of SAP strategies now are beginning to start paying off. I mean, we've talked about this notion of this Intelligent Enterprise. What is that? That's a sustainable enterprise that allows you to be able to manage things like ESG data, a green balance sheet along with, of course, your cash balance sheet, the ability to now start infusing a very significant platform, which we call the Business Technology Platform or BTP, alongside that, and of course, this substantial wave, of course, around GenAI and, of course, AI in general. If you add those pieces together, SAP is exceedingly well positioned to be able to take a significant position within the market.
Yeah. What are the pieces that you're adding together? You had GenAI.
GenAI, BTP.
BTP.
BTP.
What was the third thing?
Of course, you've got this whole innovation around the Intelligent Enterprise. This is the evolution of SAP's substantial SaaS application stack.
Yeah. Where do you feel we are on the transition in terms of getting your customers to SAP S/4HANA at this point?
I mean, look, we've got a substantial number of customers that are already engaged on the SAP S/4HANA journey. So that's maybe the step number one. What does that mean? Well, it means that you take your historical landscape, your ERP landscape, and you modernize that. You transform it into something which is a lot more capable of giving you those predictive capabilities of forecasting, all the other things that you'd be looking for from core business applications. So that's well and truly underway. The RISE motion of SAP, which is really the opportunity to take a more of a choreographed move, using SAP to manage and facilitate that, working alongside with its ecosystem of partners, and allowing the multitude of customers to be able to take that journey in a more of a hosted fashion.
So this is more or less SAP takes on the responsibility of the state of the customer and then allows them, whilst they transition, to be able to also apply the Business Technology Platform, let's call it transformation as a service, to be able to really start extending and enriching the use cases. So those are the really key moving parts of the strategy there.
Maybe you'll walk us through now. I'm going off-script, but it's interesting. So walk us through an example, if you can, of that RISE with SAP motion right?
Yeah. So if you imagine a very large, I don't know, automotive company or a large pharmaceutical company, I mean, SAP has a substantial footprint. I mean, we kind of pride ourselves, but we pride our customers more importantly. 80% of the world's business transactions run on an SAP application stack. And that could be our core financials. It could be running on the manufacturing, supply chain side. So huge levels of customers are running across a multitude of different industries. So what they typically have done over the years, and I think this has been both a blessing and a curse of SAP, SAP has been selling standard applications for specific industries that become heavily customized, right?
Because if you imagine that you have a business process, whatever that business process may be, order to cash or whether it's hire to retire, these processes are created in a foundation which is hugely beneficial to customers as a starting point. But then you get into the whole of the challenges of most organizations, different geographies, different sort of jurisdictions, et cetera. You tend to have different augmentations to those processes. So therefore, you end up with a heavily customized foundation. Now, that is not necessarily conducive to SaaS applications because you need a clean core if you're wanting to run in a SaaS environment. You think about a classical SaaS application today, we typically would patch that. You'd end up with revisions, enhancements, which we delivered on a quarterly cycle or sometimes even quicker.
Therefore, the SAP core in itself, heavily customized, very, very substantial but sophisticated set of business processes, RISE in itself presents a significant opportunity where we take on board the ownership and the accountability of running the estate, but then we also, by wrapping around the Business Technology Platform, give you a mechanism of being able to transform the core, clean the core, and then extend it and, of course, enrich it. So large customers will take that first journey, take SAP's opportunity of simplifying the estate, and then really using the BTP, which in itself is rising up. We started reporting on the BTP business, which in itself is becoming a multi-billion euro business for SAP now. And it's a combination of different solution areas. So maybe just a quick snapshot view of the Business Technology Platform.
Now, today, if you look at any of the cloud providers, the hyperscalers, they all have their own sophisticated foundation or different services. Azure, Microsoft, of course, has, I mean, Azure in itself is, by definition, it's an ever-evolving set of services. And those services will be in data management, in analytics, now, of course, in GenAI and the rest. So SAP has co-located its Business Technology Platform alongside any and all of the different hyperscalers. So if you think about the three North American hyperscalers, we have our Business Technology Platform, which is comprised of really four key subsolution areas. So you think about analytics and, of course, data management. Think about the whole of the application development and integration. And then you think about GenAI. These are areas and categories of different services that we have in our Business Technology Platform.
In aggregate, we have about 100 services that run within the Business Technology Platform. They're used to be able to extend the value of SAP customers. By doing so, running it in the infrastructure of the hyperscalers, it's not about us having to say, "Well, you can only get those services if you come to an SAP cloud." We actually have a multi-cloud strategy from day one, and we deploy these Business Technology Platform services across the entirety of all these different hyperscalers.
There's 100 services, and there were four categories. You had analytics, GenAI, what were the other two?
Data management.
Data management.
Analytics, and of course, the whole of the application development.
Application, yeah.
You're writing a research now, are you?
No. I mean, look, I've been covering this company forever. In this part of the story, I'm not where I should be on it. It's super interesting, right? It's super interesting. So how does the Business Technology Platform and RISE with SAP motion work together?
Yeah. So that's a great question because first and foremost, as I said, standard software, heavily customized. You need to get to some kind of a clean core. And what is a clean core? Well, it means that ultimately, at some stage, and we have a cloud ERP movement as well now, where this is more of a classical, I can't say classical, more of a modern SaaS stack, which runs completely independently, not managed by the customer, right? It's an SAP-managed service.
So you've got a customer-managed tenant, which is what RISE is, and you've got an SAP-managed tenant, which effectively is what the public cloud SaaS offering would then be. To bridge the gap between the two, some customers may elect to go completely greenfield, go straight to a public SaaS offering managed by the vendor, in this case, SAP, or they choose to actually not forgo all of the significant value that they have within their ERPs over the last several decades or years and want to make sure that they can continue to extend that. This is where the Business Technology Platform is a close sibling to all of the customers that actually have a RISE motion with SAP and independently, right? BTP is a standalone set of services. They run with SAP and independent of SAP data.
But the benefit to a RISE customer is that if you wanted to have a very significant level of analytics that is being built upon SAP, where we embed certain levels of analytics, I'll give you some of the specific services like SAP's Analytics Cloud. This is a differentiator for SAP customers because it will allow them to be able to start modeling, building maybe exactly from a planning perspective. For example, if you wanted to build an end-to-end planning foundation, like looking at supply chain planning, workforce planning, maybe financial planning, and you wanted to have one contiguous planning view of your in-store base, you can actually achieve that using SAP's Analytics Cloud.
The benefit of a BTP is really there to allow you to be able to accelerate the innovation on top of a core, which is going through a process of renovation and becoming a cleaner core. Then eventually, you'll land in a public SaaS environment if you choose to do so as well. Completely obfuscating your need to have to manage a private tenant, you can actually have a public managed service by SAP.
Wow. And what part of the AI solutions within BTP are people the most excited about? Is it the Analytics Cloud that you were just talking about?
Yeah. I mean, AI, for the most part, has created huge opportunity. But just as we've seen the high level of hallucinations that take place if you have LLMs or even foundational models that you create, but it doesn't have the right context. So there's two elements that I would say that's really sitting well with customers right now. The one is SAP announced something called Joule. So this is the SAP copilot. So this gives you the ability to start asking like a prompt-based kind of request in 40+ languages. You can ask any specific type of question and then know that you've got all of this foundational information from SAP applications, excellence in business process, going across 20+ industries. That is the domain of value of SAP.
If you're able to infuse that directly with an AI based on applications embedded within the SaaS applications of SAP, but on top of which, take, for example, SAP Analytics Cloud, and you want to say, "Well, as a planner, build me a model that looks for productivity for 10% efficiency gains but not diluting my margins by more than 5%." So you can start doing all of those type of what-if scenarios and using the foundations of AI plus, of course, the technology stack and the know-how of SAP's data capabilities to really be able to radically accelerate those type of innovations.
You got to say that example again. Okay. So what was it?
You're a planner. Let's assume you're a planner.
Yeah.
You're now sitting behind the planning console. Typically, you'd end up having to create multiple models. You need to create a financial model. You need to create, for example, a model that's going to be able to talk to you about your workforce and knowing, for example, how many FTEs you need versus contingent labor that you may need.
Yeah.
So now having a simple, unified experience now where you could use the Joule-like experience to ask a question, which a planner ordinarily would need to spend a significant amount of time curating all of the different data, all of the different models, creating all of the different sort of information that you need at your disposal, now it's much more driven by a question and an answer approach. And we use things like knowledge graphs, and we, of course, we're adding additional technology components. But in simplistic terms, it's really using natural language query across multiple languages, like 40 languages that we support, localized, and making it very, very easy for you to ask those very typically open-ended type of questions, which are normally very complex and very difficult to model if you go through a traditional kind of an approach that you would have to take here.
How do you answer it? So I get the asking the question part, right?
Yeah. The way
How do you answer it?
How does it answer it?
Yeah. I mean, the way it answers it is by connecting all of the different repositories and sometimes silos of data that you need to connect together. And this kind of takes us on maybe slightly, I f I may indulge you here a little bit, on the broader data management challenges that most customers are faced with and where the current market dynamics are not necessarily supporting all those requirements. Now, on the one hand, I mean, the framing of the conversation is a lakehouse architecture. Most customers will know what a lakehouse architecture is. It typically is taking a substantial repository, let's call it a data lake, now more or less a cloud data lake. And then you have a warehouse on top of that. This is now where it becomes a lakehouse because you're combining both pieces together. That in itself is a value add.
But then the assumption is that you have to have a substantial amount of data movement to locate the data in these lakehouses. And that, by definition, is a body of work that some customers are either ahead of or others are trying to reconcile how much can they put into that. So assume that you've now I'll give you the practical analogy here. O f course, you've got a busy lifestyle. And let's assume you're at home and you've got a dedicated day, Sunday, that you dedicate to doing all your filing, all right? And like me, if you're like me, I'm just all over the place, right? I've got data that sits across multiple areas in my house. I've got my filing cabinet in my office. I've got data that sits, of course, in multiple parts of different parts of the house.
You can imagine the enterprise almost resembling that. Data is filed away in multiple different environments. If my answer to you was on a Sunday, what you need to do is go and buy yourself a big-ass filing cabinet. Go to Staples, get yourself a big filing cabinet, and go around all of these different repositories that you have, collect all the paper, loosely coupled, file it, organize it, and then store it in this big filing cabinet. Some would say, "Yeah, that's a great idea. It makes a lot of sense. And I can sense the value in doing that because I'll have one significant, authoritative store where all my data resides." But the reality is, the moment you've done that, you're going to suddenly find yourself creating more filing cabinets elsewhere as well.
So then you've got to start creating the whole of the ELT, the extract, the landing, and the transformational part, the pipelining, the data pipelining, understanding that the governance and, of course, the whole of the foundation of data in itself is constantly evolving, and it's highly distributed. So SAP's response to that has been around creating what we call the SAP business data fabric. And the fabric, by definition, is not a physical. It's a virtual representation. And we have a service called Datasphere. We have an Unleashed event taking place this week, in fact, where I'll ask you all to maybe tune in and listen to that. The notion of the Datasphere is really the embodiment of a business virtual fabric.
It allows you to connect to all of those sources, being able to semantically onboard the information that you need, and have a simplified way of being able to do that modeling exercise that I described because you don't need to go and figure out where are the physical data sets by semantically modeling and creating one semantical representation, be it virtual, in the Datasphere, SAP's business data fabric architecture. That means that you can then accelerate the delivery of those type of use cases. So that's the how part.
You know, this whole no copy is basically the same.
Yeah. I mean, the no copy fundamentals are that why would you keep creating multiple shadow copies of the same data? And then, by definition, the deterioration of the value of that data because you have to somehow have coherency back to the source. And this is primarily the big problem that most customers have been faced with for several decades.
Yeah. This is a big trend, right? Because Salesforce is talking about the same thing.
Yeah. I mean, Salesforce, arguably.
The whole idea of the CDP seems to be diminishing. Is that too strong a statement?
So what is, I mean, Salesforce, you bring it up. It's a good example. Salesforce announced the Data Cloud, right? I think in their earnings transcript, they talked heavily about the value of the Data Cloud going forward. Now, what does that give them in relation to what they had in the past? So they had a movement, which was to try to bring data under one Customer Data Platform. And they chose to do so because it was combining Commerce, Marketing, Sales Cloud data. But by doing so, that creates a movement. That's the filing cabinet movement, right? So getting data under management is not trivial. But then they also realized that there is this big movement that most customers have already selected a platform of their own choice. That could be a Databricks, a Snowflake, a BigQuery, whomever that person or vendor is.
So the governance of that data then has to assume that data is predominantly going to be sitting in these environments. And that's also another. I'll be very, very transparent from an SAP perspective, in our business data fabric discussions and in our Datasphere delivery, we took an open approach with the ecosystem. We said, "Dear customer, if you've elected to land your data in a Databricks environment in the Delta Lake, fine. We support that. And Datasphere will have the mechanism of being able to load data semantically enriched, so not just raw data where you lose the semantical value. And we can land the data directly in the direction of those chosen platforms of the customer." In fact, I'll be bold enough to say that it's a customer stack. It's never an SAP stack. It's never a Microsoft stack.
It's never an Oracle or whomever vendor stack. The quicker we recognize that and the more open we are in that approach, the better it will be for the end customer.
That's great. We have two and a half minutes left. Do we have any questions from our audience?
Yeah. So Datasphere is still relatively new. It's kind of the next generation of Data Warehouse Cloud. So what are you seeing in terms of demand there? And also, it would be helpful to walk us through some of the partnerships. I mean, you have partnerships with Confluent, DataRobot, Databricks, Collibra, some of the companies that are presenting here. So that would be interesting to hear about.
Certainly. Thank you for the question. So firstly, Datasphere has been around just under a year. In fact, it'll be a year celebration actually this month. But it's a successor, as you mentioned, to Data Warehouse Cloud, which is a product that's been around for three or four years now. We have over 1,300+ customers already now on the Datasphere because it was a succession of movement of those customers' landscapes over to Datasphere. And the notion of having an open ecosystem approach has really given us the key value, as I alluded to here. So when a customer has elected to have Delta Lake and maybe using the open table format of Delta or Databricks, rather, we support that natively.
Therefore, take, for instance, a large, I don't know, a retailer, for example, where ordinarily, if you're having to just take a raw extract out of SAP, load that into Databricks and into their Delta Share or Delta Lake, you're losing, as I said, that fundamental value, which is a semantical layer. If you lose that, a good example of this is we talk about things now in data products. An example of a data product would be an invoice, right? Think of an invoice as being a concatenation of multiple different data sets, right? You've got supplier information, order information. There could be tax information. Raw data, if you were to extract it, would be the individual silos of the data that I just described.
So therefore, having a semantically enriched data product and being able to land it inside of a data lake like Databricks, for example, can give anything to 30%-40% productivity value immediately to an end customer because they're not having to reconstitute or recreate the entire semantical layer. Confluent is another great example where you have Kafka now, which is a managed layer of message transportation and delivery. We see many customers embracing Kafka as that messaging transport layer. So us having an ability to transport those data products, enriched and, of course, given value add in the kind of the Confluent environment, also gives a tremendous value add. And we've also announced the extension of our partnership along with Collibra, Collibra being, of course, privacy, cataloging, lineage.
When we look at the whole governance around AI and the capabilities that they now have to be able to do ethical AI means you need a governance layer. So we're even doing a deeper level of partnership with the likes of Collibra as well. So really using the ecosystem not as a nice-to-have and just to show in optics and looks great on paper, but this is really, from an engineering perspective, putting that level of development work together so we can land those capabilities natively within the stacks of those partners.
Irfan, thank you so much. We went over. I got through like a third of my questions. So it was a real treat for us to have you here. There's a lot for us to learn. We really appreciate it.
Thank you very much. Thank you for the opportunity.
I really appreciate it.
Thank you.