How to Get Real-Time SAP Insights With Databricks SQL

Summary

• SAP Business Data Cloud Partnership: Databricks and SAP announced a partnership to simplify access to SAP business data through Databricks Data and AI platform.
• Query Business Data Without ETL: The integration enables instant scale, automatic optimizations, and built-in governance across all enterprise analytics data.
• Real-World Implementation: Zalando demonstrates how to leverage this integration for fast, AI-powered insights from SAP data with Genie Spaces.

How to Get Real-Time SAP Insights With Databricks SQL

Talk by Alejandro Saucedo, Director of Eng, Science & Product, Zalando SE ; Jon Levine (JPL), Product Manager, Databricks ; Olaf Melchior, Director of Data Platform, Zalando SE
Tired of waiting on SAP data? Join this video to see how Databricks and SAP make it easy to query business-ready data—no ETL. With Databricks SQL, you’ll get instant scale, automatic optimizations, and built-in governance across all your enterprise analytics data. Fast and AI-powered insights from SAP data are finally possible—and this is how.
Intelligent analytics for everyone: https://www.databricks.com/product/business-intelligence
BI meets AI: https://www.databricks.com/resources/ebook/business-intelligence-meets-ai
Explore the only unified platform for agent systems: https://www.databricks.com/product/artificial-intelligence
See all the product announcements from Data + AI Summit: https://www.databricks.com/events/dataaisummit-2025-announcements

Chapters

• 0:00 Intro
• 1:35 SAP Business Data Cloud
• 6:28 Challenges
• 14:57 AI Progress
• 16:24 AIBI
• 18:07 Genie Spaces
• 19:58 Databricks Journey
• 22:12 Use Cases
• 27:39 Takeaways

FAQs

What is SAP Business Data Cloud?

SAP Business Data Cloud (BDC) is a platform for all of your business data in SAP, consolidating data from various SAP applications like S4 HANA and SuccessFactors.

Why is the Databricks and SAP partnership important?

The partnership makes it easy to bring important business data that customers have in SAP into Databricks Data and AI platform without complicated ETL processes.

What company provides real-world insights in this video?

Zalando, a major e-commerce company, demonstrates how they use Databricks SQL and Genie Spaces to get real-time SAP insights.

What are the key benefits of this integration?

Benefits include instant scale, automatic optimizations, built-in governance across analytics data, and AI-powered insights through Genie Spaces.

Full transcript

0:00 Intro

All right folks, good morning. How's everybody feeling? All right, welcome to day two of data and AI summit. So we have our one of the session today with uh data bricks on SAP and for that I would welcome who is one of our accomplished PM in data bricks. Please give a big round of applause for him. Okay, thank you. And then we have from Zalando. They're gonna they're going to come up halfway through. Yeah, we'll give them a warm introduction when they're here.
Awesome. All right. Okay. I know it's almost 8 a. m. Thanks everyone for uh joining us bright and early uh data summit. So um Oh, do you have the clicker? I do. Got the clicker. All right. So um quick reminder, there are some forward-looking statements here. Um, please do fill out the surveys. We'd love to hear from all of you. Um, so I'm here to tell you about uh data bricks and SAP. So since you're all here at data and AI summit, you probably know a bit about data bricks, we don't need to introduce data bricks.
And I'm guessing if you're in an SAP session uh here at 8 a. m. on Tuesday, um you're familiar with SAP as well. So, I won't give too much of an introduction, but um many of you probably know we introduced we announced uh earlier this year a partnership um between our two companies. And we're really excited about this partnership. Um in short the reason we did this is because we wanted to take some of the most important business data which our customers have which is in SAP and we wanted to make it easy to bring into what we think is the best data platform in the world which is data bricks.

1:35 SAP Business Data Cloud

Um so before we get we're going to talk in this presentation a lot about like why we did this. Um we're going to hear from one of our customers Zolando. I want to leave lots of time for everyone here to um I'm sure you all have lots of questions but I just want to give a sort of quick high level overview about um what actually is this partnership. What did we actually announce? So SAP first of all has introduced something called business data cloud or BDC and BDC is a platform for all of your business data in SAP.
So SAP as many of you know has a lot of applications. So there's um you know the sort of core ERP application which is like you know S4 HANA um there's you know success factors arba and all of these applications are producing data and BDC business data cloud is that platform for all of your SAP data. Now BDC contains a number of things. Um one thing that's really important to understand are the data products. So data products are curated managed data sets which are produced by SAP and you can actually think of these as like u managed pipelines that in data bricks parlament they create like gold tables.
So they're like nicely um nicely consumable nicely usable tables that you can then use throughout BDC. Now for the data bricks side of this there's really two things that we're doing as part of this integration. So the first one is that we have a new version of data bricks. Um think of this as like an OEM version of data bicks. It's included as an application inside of BDC and um it's a managed application. So what that means is SAP um you know they help you run it.
They control you know the identity all the data is stored in SAP managed storage and it's really tightly integrated with BDC and that's something that you get if you purchase BDC. And our intent there is really for folks who are new to data bricks, maybe who aren't using data bricks yet as their enterprise data platform that's a really great place for them to get started with data bicks. Now I'm guessing most of the folks in the audience uh we'll get to we'll do a quick show of hands in a minute.
I'm guessing most of the folks here are using data bricks today. If you are an existing data bricks customer, our integration is also going to allow you to delta share these data products from business data cloud into your existing instance of data bricks and that's what we're going to talk a bit more about today. That's what we're going to focus on. Um just really quick so everyone know not to you know spoiler alert um the OEM version SAP data bricks is actually live today.
If you're an SAP business data cloud customer it's live on AWS. were generally available. The integration, the delta sharing integration with the standalone version of data bricks coming very soon. Okay, so why do we do this? What can you actually do? Um, well, there's so many use cases which I'm excited to talk to you about. We're going to talk about all these. We're going to hear from Zando about what they're doing with SAP.
But in this talk, we're actually going to focus on analytics. Um, and the idea and this idea of interacting with your data with natural language. Um, but I think what's cool about this integration is there just, you know, a lot of our customers have done this work already. And there's so many amazing things they're doing. So when I talk to finance teams, um, you know, they're running financial forecasts to do that. They need not just their SAP data, but they actually need data which is in all these other systems across the enterprise.
They're running planning exercises. Um, we hear from teams that are using data bricks to help them close their books faster at the end of the month. We hear from customers who are building customer 360 tools, right? It's very common to ingest that data from your CRM, but it's much more powerful if you actually have the order history, the transactional data, which really often lives in SAP. And for folks who are building AI, the SAP data is really critical.
So, um, we've heard from manufacturing companies, uh, mining companies that are doing predictive maintenance on their equipment. So, they're actually able to predict when their equipment is going to fail. And the really key operational data frequently lives in SAP. they're able to use that data to train AI models and build these forecasts. So, um, just talked about all these amazing things that our customers are doing.
If you know we have this amazing partnership today, what why did we need all this stuff we just announced? Well, the reality is that the majority of data in SAP actually remains untapped for AI. So, let's do a quick show of hands. How many folks here are data bicks customers today using data bricks in some capacity? Okay, most people not everyone. And then how many people have already keep your hand up if you are bringing SAP data into data bicks?
Oh, okay. Actually, good amount of folks here. Okay, so how many people would recommend that? 10 out of 10. That was like a fun experience. You did it again. It was awesome. Keep your hand up. Yeah, not as many people. Yeah. Um, it's really hard. It's been really challenging to bring SAP data into data bricks and that is what we wanted our partnership to solve for those of you who've already done it. Um, amazing. We're going to talk more about what you can do.

6:28 Challenges

Hopefully, we're going to make that easier is to maintain that to keep that going. For everyone else, we want to make that easy to bring that data in. Um, I was actually I was just at uh, which is the SAP conference in. It's amazing how many teams we talked to that are SAP teams. They're like, "Yeah, there's this like there's this like enterprise data platform we have over here and we don't talk to them. We just live in our own world."
So, we want to fix that. So, what are the challenges that SAP customers have? Well, there's four challenges. These are not necessarily specific to SAP, but I think they're really acute for these for our joint customers. We're going to talk about each of these, but the first one is scale, fragmentation, governance, and AI. So, let's talk about each of them. So um this scale is really one of the original challenges that data bricks helped solve.
Um so the problem is that traditional warehousing solutions costs just grow out of control. They just grow exponentially with data size. And the core problem is that if you're bundling storage and compute as your data sizes grow you're paying for all this compute which you're not using most of the time. And I think most most folks know this if you're data bricks customers. This is one of the core things we've solved with the lakehouse architecture.
We've separated out the compute and storage. So it's not uncommon that I hear from an SAP customer like oh I have a you know I have a warehouse with one terabyte or 10 terabytes of data and that's a huge problem. Of course at data bricks we have customers with hundreds of terabytes pabytes xabytes of data and they're able to do that because they're keeping their data in open formats in commodity storage. It's very cheap to store the data and you pay for the compute you use.
And so at data bricks, you know, we say may the best engine win. And um of course we think we have a very good engine as well. And so I won't I won't talk through all these graphs in detail, but across like a range of um you know, data sizes and price points, we think we have really competitive price performance. But again, the idea of the lakehouse is that we've kind of broken that cost curve by saying you can store as much data as you want and you're paying separately for compute.
So that's that's just a really core foundational idea for data bricks. Now I mentioned cost um of course scale is a huge driver of cost right if you have a lot of data that's and you're you know quering it a lot that's going to cost a lot of money but also cost is just driven by complexity and by fragmentation and so again just in general data brick sling we hear from customers is the data landscape is so complex um there's so many things they need to do from you know ingestion um data governance machine learning and they use a lot of different vendors for this and very frequently people are like oh yeah I have like one of every type of vendor for you know many of every type of vendor.
It's not uncommon to hear that. Um and for SAP in particular, it's very common that we hear folks have sort of a patchwork of tools and integrations just to migrate the data from SAP to data bricks. Um and it's just really hard to do that. So just want to talk a little bit more detail about sort of the current state of you know how people are people are doing that migration. So SAP in general requires a license to migrate the data.
So you typically need like a runtime license or something called premium out integration and then people often use like a vendor to do that and you know the vendors obviously have to pay the vendor itself and like the licensing terms can change um you actually have to even though you're you're paying a vendor or sometimes you know an SI you still have to think about the load on your source systems right because they're actually running queries on those source systems all the time and those are business critical so how do you make sure those don't fall over um and then there's actual like data engineering work so you have to model that data data, you have to like maintain this integration.
And really like none of the things that I just mentioned, none of that's actually adding value to your business. All of that is just overhead that you need to do just to be able to unsilo that data to bring that SAP data into the platform. And so what what actually happens in practice when we talk to customers is that's just not worth it. They don't end up paying that overhead cost and the SAP data stays separate. Those teams live separately. those functions are just working in a in a separate environment.
So um how do we solve this? Well, there's a number of things we're doing. We want to take a fundamentally different approach with our partnership. And um you know the first one is just we're trying to eliminate a lot of the complexity around licensing and vendors. So this is like this is the golden path. This is this is what we recommend like you know we're telling everyone this is how you get your SAP data into data bricks.
Um I'll just mention briefly because I know a lot of people ask about this. when you purchase a data product in SAP business data cloud, you actually get a license to share that data into data bricks. So, we're just kind of solving this problem of how the integration actually works. Licensing questions, you know, need for other vendors, the integration just works. We're responsible for that. Um the second thing which I think is really important is that SAP through the data products is actually taking on a lot of the complexity around modeling the data thinking about the load on your source systems.
That actually becomes SAP's responsibility with the standard data products. One thing I'll just uh shout out really briefly, I know a lot of folks here are probably doing some of their own data modeling and tools like you know BW or Datasphere today. You will actually be able to share those as as custom data products as well. But we think a lot of folks like they just haven't even taken on that challenge and the standard data proc products are going to solve a lot of those headaches.
And then finally I mentioned on the data engineering side there's all these things you have to worry about like you know what type of CDC am I using or um uh you know is the data drifting between SAP and my analytics platform and you know are there delays in the data like we actually hear from customers who say oh yeah I just pulled SAP data in you know it's like once a week once a month. Um with delta sharing we solve all that.
So I didn't talk much about delta sharing yet but the essence of delta sharing is that it's it's zero copy sharing. So these data products are written into you know again in an open format into SAP managed storage and with a delta sharing protocol which is an open source protocol. Data bicks is able to read and query that data and govern that data through unity catalog in a way that it just looks like it's it's your data right so it's just super easy to read it.
As soon as that data product gets updated, you're able to then query that data and use that data. Okay, so I briefly touched on governance. I think we all know governance is a huge challenge. Um, you know, security and privacy are under threat. There's, um, data regulation all around the world. There's new regulations happening, you know, every month. Um, cyber attacks are on the rise against companies against data platforms.
And SAP data is very often the most sensitive data in the enterprise, right? Um, very often it's regulated. There's some compliance standard or multiple compliance standards you have to meet. Typically contains PII. Companies typically have retention policies. So you have to keep the data for a certain amount of time or you can't keep it past a certain time horizon. There's very complex permission models. So the application owners and these source systems are actually modeling the rules of like who can use and who can access this data.
So what happens when you bring that data into your analytics platform? What happens when you start to derive downstream data sets from that? Um what happens if you train an AI model on that data? What governs that then? So um of course as many of you know our solution to this at data bricks is unity catalog. So unity catalog provides comprehensive governance for the lakehouse. Again the idea of the lakehouse is that you're keeping all this data um in open formats right in commodity storage.
So you need a single view to actually manage and govern that data. Um and Unity catalog as we show here, it's not just your structured data, right? It's your unstructured data. It's the AI models. And it's way more than just access control, right? Because there's things like row level security. We have metric views which give you that semantic layer. You can actually see the lineage, understand what is the data model, where your where that data actually comes from.
So UC is giving you the tools to actually govern that data and make sure it stays compliant when it comes to the analytics platform. So I did I did touch very briefly. I mentioned that you know UC helps you govern AI models. I think unless you've been under a rock um I think you're going to hear a lot of this week about the amazing progress in AI. This chart is actually a year old but what this is just showing is that um the models are getting better and better. you know, every every month, day, hour, uh there's a new incredible frontier model that's coming out.

14:57 AI Progress

Um even the, you know, open source models are incredible. They're so good that the benchmarks we use to measure them are getting saturated, which is pretty amazing. And some people say we may be nearing artificial general intelligence, which is pretty incredible milestone, except that even if you had a truly generally intelligent model, it wouldn't be very useful necessarily. And I think this customer put it really well.
They said, you know, I don't care about the standard benchmarks. I just want the model to do well and understand my data. So that's why data bricks, we care so much about data intelligence. What that means is that the AI models, they actually have to understand your business, your customers, your data. If they don't do that, um, they're not very useful. So this week, um, you're going to hear a ton about how data bricks is helping you to build AI systems that are really intelligent.
And so there's a whole suite of things we're doing. I'm not going to get into too much detail, but just to quickly mention them, um, you know, what do you actually need to do to make to help build one of these AI systems? Well, um, uh, you need to make the data accessible. So, obviously, you know, need to do all the data engineering we just talked about. Um, you need to expose it in a way that the models can use. So, maybe a vector index, you know, feature store.
You need to actually be able to train the model, right? Whether that's fine-tuning or training a model from scratch. You need to be able to deploy it. You need to understand how that model is working in production. You of course need to govern it with Unity catalog as we mentioned. So there's gonna be some great talks throughout the week on that. In fact later today I think there's a talk specifically on um uh building AI with SAP data I think later today.

16:24 AIBI

Um but what I actually want to talk about is that AI is really powerful for helping with analytics. And so many of you know we have um our offering we call AIBI which is intelligent analytics for everyone. And we think there's really two important things that AIBI helps with. So the first is for data practitioners. And I expect most of the folks here in the room or here at the um here at the conference today, you're either a data practitioner, maybe you're maybe you're managing data practitioners here.
We think AI is a really powerful tool to help you in your job. So to help you write SQL, to help you prepare the dashboards, um to help you do things like manage the underlying compute, right? We don't think that practitioners should be spending time, you know, like tuning knobs in their SQL engine um, you know, figuring out like, you know, the optimal indices and things like that. AI can help you much more quickly actually provide value off that data.
So that's the that's the top half. The second half is that we think that AI actually gives us, you know, kind of this holy grail of actually letting non-technical folks, folks who are not running SQL, folks who maybe don't have a deep understanding of the data model, actually interact with that data. And our belief is that you know if someone's like a nontechnical user they're not just going to like go to chatgbt and ask it to like write a SQL query for them.
The problem right is that it doesn't understand it doesn't understand the underlying data. So what we want to do is give the practitioners that's again folks here in this room or again folks folks you manage give them the tools to be able to curate that data. It's something we have um called Genie Spaces which actually let this nontechnical user um uh talk about this data. So we're going to have some great talks this week on AIBI.
You're going to hear about some great things we're doing in the keynote. Um, uh, but rather than hear me talk more about it, I'm very excited, uh, to welcome up our customers, Alando. So, yeah,, please come up. Thank you guys. Awesome. Um, so these fine folks, we I can I can stand on this side of the podium. All right, there we go. um they're they've actually done the hard work to bring SAP data into data bricks and uh we're just going to hear more from them about how they've actually done that and what they've done.

18:07 Genie Spaces

So yeah, do you want to introduce yourselves quickly? Sure. So yeah, nice to be here. Um good morning. So I'm the director of Zelando data foundation. I'm for 12 years with Zelando and I'm working with the team that is responsible for basically the whole data analytics foundational platform including governance, business intelligence, etc. And I'm not really working with SAP. I also have to say this, but I'm actually the peer of my co- so the director of ERP.
So and uh I have witnessed and supported the work to get SAP data to data fabrics in the past. Awesome. Great. And yeah, hello uh good morning everyone. Uh. I am director of engineering uh science and product at Salando. So I'm responsible for one of the uh central data owning uh platforms. So basically we build a lot of the data and AI on top of those central infrastructure. So we have a lot of use cases that power supply and demand uh for Salando including predicting demand predicting supply.
So a lot of relevant things that we're going to cover today forecasting marketing intelligence and um maybe a little bit about Zando um um so some of you may not know about Zando as we are primarily based in Europe. So we are one of the leading uh fashion e-commerce u organizations in Europe. We have about 52 million customers uh 20,000 employees uh about 16 billion uh GMV. Uh so we require basically this data and AI foundation to operate at that at that scale and that's what we're going to be talking about today on the transactional data.

19:58 Databricks Journey

Amazing. Yeah. Maybe just really briefly tell us about your journey with data bricks and then your journey of like bringing SAP data into data bricks. How did that come about? Yeah. So, um, when I joined Zelando, which was in 2012, so Zelando was founded in 28 when when our two founders just started shipping flip-flops from a flat in Berlin. And so um the transition we have made is basically one from let's say a low scale low complexity data landscape where we basically just had we had a transactional monolith we had a um monolithic data warehouse based on uh a well-known company that we don't work with anymore but so the and um basically in when when we entered hypers scale we basically saw the big need to rearchitect the whole technology landscape which meant we went went to microservices that ISOly communicate back Kafka etc.
So we just resolved all the monoliths and then we had of course this big um data integration challenge that came with that because monoliths are easy in the sense of that you just have one source when you have thousands of event streams that you actually have to integrate you need something else. This is where when where when we actually moved to introducing our data lake in 2016 which worked okayish in the beginning but then also we quickly saw we need a partner.
We found this partner in in data bricks I think in 2018 and so and basically since then we are somehow in the business of resolving all data silos that we can find in the company and there are still many left and one of them we are talking about today. Yeah maybe let's let's talk a little bit more about SAP. So um I think one thing we're excited to share is just for folks who've actually done this journey what does it actually look like?
Maybe you can tell us about some of the use cases you have on data bricks um that involve SAP data. I know it's it's usually SAP data plus other data but yeah I would love to hear about some of those use cases. Yeah. Yeah. So so let's talk about the use cases and I think it'd be also good to cover about like the integration learnings. Amazing. Yeah. Um but indeed so one of the flagship use cases that is also worth covering here is uh what we refer to internally as our autonomous forecasting system.

22:12 Use Cases

So we basically were able to build on top of transactional data um what is deep learning models that hoover a lot of uh kind of like insights from our market combining also you know macroeconomic insights uh to be able to capture signal that allows us to predict demand. So we're using this at different granularities and levels different use cases. So you can imagine things like capacity planning for our logistics also for topline setting for our P&L as well as many other things and what is interesting as well is that we have started also leveraging foundation models for time series which is actually like an emerging trend.
Um so we have been able to just you know not only leverage what are those zero sherot type kind of approaches where where you have a model that has not yet seen the data and is able to do pretty good but we were able to actually like leverage them in for example new markets where we're expanding into where we don't have historical data and as we then capture that through our systems integrations we can then feed it in.
Yeah that's amazing. Yeah Olaf we were going to mention another use case. Yeah, I think this is this is already a good list. I think there so so this whole forecasting space also profit contributions. So so there are all the there are obviously quite a number of use cases where you just actually want to define the the financial perspective on the quantitative perspective that you have in the transactional systems and so one of them is profit contribution.
So for instance, so how we steer the business, I think this is not super original, right? So, so we we have a very strong focus on on it will profit contributions per per article per market per all levels and then of course also per customers with the customer lifetime value and this is a this is a I'd say one of the northstar metrics we use in the company and that of course always requires actually combining the financial perspective with the transactional perspective and so this is one of the very I think you were telling me you had this great anecdote about you have some system that understands how many units were sold and another that understands the price so without understanding sort of the units and the price, it's actually hard to have that view.
Yeah. Exactly. And so, so this is a constant exchange basically. So, so for instance, imagine we have 13 different um different um fulfillment centers across Europe and of course items constantly move. So there's of course Yeah. So we we predict at which in which market actually this item is going to be so sold next. So it moves it's on a truck somewhere. At the same time, we of course want to know actually what is the value of that item that is currently sitting on that truck.
And so this is where an example of where these two sides have to come together. It's funny as we were preparing, I think we were talking about some of the use cases for SAP data. I think you one of you mentioned like there's almost no use case for SAP data that doesn't involve integrating it with some other data in the platform. Pretty much all these use cases are about bringing the data together, which is which is amazing.
Um, so yeah, you were starting to say maybe let's hear more about that journey to get that SAP data integrated. What did that look like? what are some of the challenges you had to overcome to do that? Yeah. Yeah. I think this was always like um people's so individual pipeline building right. So, so starting from um yeah when where you go so so what this is not firstand knowledge but alo my understanding is actually what you have originally at SAP HANA is is tables that are really really hard to understand with some cryptic names and uh so so you actually need people with with a deep understanding of actually how SAP works internally with a deep understanding of the techn technicalities of the contents and then also of the the data formats that are used on the platform. form.
So, so basically you need Yeah. So, a very exist that needs that has a list. So, so we we try to to have these people but it's of course it's a long journey right so and um people have done this and I think this has been super successful but it's a very very special profile you need people who are deep into SAP deep into the contents but also deep into data bricks and modern data stacks and so right that's just hard and this is what I think is that's cool changing yeah um I think one thing that's interesting is you two have slightly different roles and you mentioned you all have a counterpart who works in the ERP systems we we talked to at.
And I guess I'd love to hear more about like how do your teams work together and how do you how do you make sure you don't end up with those silos and end up with you know how do you how do you have good collaboration across these different kinds of teams? Yeah, I think this is so important because basically this this case wouldn't have been possible if we wouldn't be working very closely. So, so we actually we sitting very close even physically when we in the office and it's it's um let's say one of the strategically one of the big goals that we have to actually break down silos not only in technologies but actually in in in the community in the organization and so so traditionally of course this is a very different profile with very different skills etc but we al also saw seen that there's quite a lot of interest um especially so the SAP ERP engineers are actually very interested in learning about modern data technologies.
So this is also something that is that's great and if you give them the opportunity there's a lot of willingness to actually go there but it's a long journey it's not yeah yeah and I would say also um that is something that maybe for the uh folk in this room is maybe an important takeaway. So as we develop kind of like this capability I mean at Zando yes we have a data platform data foundation we have a lot of kind of uh data mesh use cases of data where we're extracting value but we are not at the end right like if anything we're still at the early uh parts of the journey um and I think this partnership between you know central functions and let's call it like business uh specific uh you know functions um is super critical right especially when it comes to um lift and shift initiatives like you know integration with SAP but also more generally with migrations I mean we all know kind of like the pains of multi-year migrations and how to achieve that I think the partnership of you know building those bridges through I mean at Saland is more like steer codes global initiatives global projects I think we call them um helps a lot kind of like uh push that but it doesn't remove some of the the gaps that we still see like for example the fact that the pipeline lines would be specific to those particular data sets as opposed to having kind of access to everything.

27:39 Takeaways

Yeah. So I think for a lot of folks here are well it sounds like there are a lot of folks who've done that integration with SAP and data bricks but maybe for folks who are getting started is there any advice you have if you could go back and especially in light of some of the announcements the partnership that we've talked about today. Yeah. Okay. So so looking back I would say the crucial part is just really to get all these perspectives in.
You need the functional perspective. You need the technical perspectives of both systems actually and also an understanding of how platforms data platforms work in general also from a functional point of view with a with a view on the ultimate customer that you have there. And I think the with the announcement so what I find exciting about that is just that it's not only removing effort in actually so having to build a pipeline but rather it's it's it's just having it sitting there.
So, so, so in case there is a new, so if and his team have a new use case for actually building some some new forecast model, they don't have to ask engineers from multiple departments now to build a pipeline, but it's just there. And I think that's actually the what's driving or what will be driving innovation, I believe. All right. So, last question. What's on your wish list? Is there something you wish we had announced in the partnership maybe we haven't announced yet that we should go do?
Yeah. Yeah. Yeah. So we were chatting about that uh just now. So I think I think uh in a world where obviously there is this data available through magic, right? Like we just you know have have it accessible. Uh one of the trends that actually we were mentioning uh I was mentioning that is emerging right now is uh foundation models that go beyond just what we are kind of seeing today which is text, image, audio. So this is foundation models for transactional data, things like uh time series, things like tabular data.
And right now we're actually seeing a trend of what is referred to as um as as tabular foundation models. So basically these are models that are built maybe on uh graph neural network architectures that plug into your databases into your relational databases and just hoover up basically your data to be able to capture signal and be able to make kind of like insights predictive insights on something right. So, so the the value of having that data available is not only that like uh business teams um are able to create new models on on on top of this data like new forecasts but even one step beyond we would be able to actually extract insights um across the data that is available right so then we can maybe identify things that perhaps we wouldn't know that we could capture value from right so I think that would be where the I guess golden nugget would be Um and from our side this is the opportunity uh that could actually you know push uh to the next level as well.
That's awesome. Yeah. One one thing that I also find very very um very attractive about the um announcement is the role that data products play and so so that we we move away from just the technical uh data structures that are idiosyncratically existing in in the different systems but rather something that is actually useful beyond that scope and maybe so so to to wish ambitiously I would hope to see more of these right so into other enterprise platforms as well and this thinking actually saying so so we have data products we have information that is not only for our own system but rather that is actually for the enterprise and that is useful and uh well documented makes sense yeah cool all right thanks so much for joining us today thanks really appreciate your time thank you all right just going to really quickly wrap up here um so just to recap again uh SAP data bricks which is this OEM version of data bricks bundled version it's included in business data cloud if you have BD DC you can use this today.
It is generally available on AWS. Um we are bringing the ver the GCP and Azure versions very very soon and very soon we will have the delta sharing the native delta sharing integration with your existing data bricks um coming later this year. So stay tuned. Um very excited to have that in everyone's hands. Um and uh quick shout out please do we have a bunch of sessions this week about our SAP and data bricks partnership.
You're going to be hearing a lot about it. So please do check out um some of the other sessions. We're going to talk about AI, go really deep on actually Delta sharing technology itself. So encourage folks to check those out.

Learn more about the Databricks Data and AI platform.