Sponsored by: SAP | SAP and Databricks Open a Bold New Era of Data and AI
Summary
- SAP and Databricks partnership: SAP and Databricks have formed a landmark partnership combining SAP's business process expertise with Databricks' AI, machine learning and data engineering capabilities.
- Business Data Cloud integration: SAP Business Data Cloud now features Databricks as an embedded OEM product, enabling unified data and AI solutions for enterprise customers.
- Enterprise-scale AI: The partnership enables organizations to build AI solutions using trusted business data from SAP systems, powered by the Databricks Data and AI platform.
Sponsored by: SAP | SAP and Databricks Open a Bold New Era of Data and AI

Talk by H Nair, Global Head Ecosystem Integrations, BDC, SAP (Sponsor Speaker)
SAP and Databricks have formed a landmark partnership that brings together SAP's deep expertise in mission-critical business processes and semantically rich data with Databricks' industry-leading capabilities in AI, machine learning, and advanced data engineering. From curated, SAP-managed data products to zero-copy Delta Sharing integration, discover how SAP Business Data Cloud empowers data and AI professionals to build AI solutions that unlock unparalleled business insights using trusted business data.
Here’s more to explore:
Economist Impact: Unlocking enterprise AI: https://www.databricks.com/resources/webinar/unlocking-enterprise-ai
The State of Data + AI: https://www.databricks.com/resources/ebook/state-of-data-ai
See all the product announcements from Data + AI Summit: https://www.databricks.com/events/dataaisummit-2025-announcements
Chapters
0:00Partnership introduction2:15SAP strategy and vision4:36Business Data Cloud architecture11:48Databricks integration benefits13:11Impact on data engineering14:38Integrated workflow experience15:51Practical automation example16:35Data product strategy18:20Closing and wrap-up
FAQs
What is the strategic importance of the SAP and Databricks partnership?
The partnership unites SAP's deep expertise in mission-critical business processes and semantically rich data with Databricks' industry-leading AI, ML and data engineering capabilities for enterprise solutions.
How is Databricks integrated into SAP Business Data Cloud?
Databricks is now embedded as an OEM product within SAP Business Data Cloud, enabling seamless integration between SAP business data and Databricks' AI and analytics capabilities.
How can I reduce the time it takes to make new data available for both analytics and AI use cases?
Traditionally, analytics and AI have different pipelines—analytics uses batch ETL while AI needs real-time feature serving. This video explores how unified data platforms (like Databricks integrated with SAP) eliminate this duality. New data flows once into the Databricks Data and AI platform and is immediately available for both analytics queries and AI features. By consolidating pipelines, organizations reduce ETL duplication, maintenance overhead, and time-to-value. SAP's tight integration with Databricks enables customers to go from enterprise systems to analytics/AI in minutes, not weeks.
How do you support mission-critical applications and data?
Mission-critical applications require high availability, disaster recovery, data consistency, and regulatory compliance. This video shows how the Databricks Data and AI platform combined with enterprise SAP systems provides these guarantees: data is replicated across availability zones, backup/recovery processes are tested, ACID transactions prevent data corruption, and audit trails support compliance audits. For enterprises like those in the video, the reliability and governance of integrated SAP + Databricks is essential—losing data or compliance violations would be catastrophic.
Why is the enterprise data platform becoming more important to marketing strategy?
As enterprises recognize that customer data drives competitive advantage in AI, the data platform has become critical to marketing strategy. This video indicates that organizations want to use their customer data (from SAP, CRM, web analytics) to power AI agents and personalization engines. The marketing department can no longer be disconnected from data infrastructure—they need access to customer data through governance structures that protect privacy while enabling personalization. Enterprise platforms like Databricks + SAP that combine operational data with analytics and AI capabilities are now strategic for marketing.
Full transcript
0:00 Partnership introduction
You're here to talk about you're here to hear about SAP and data bricks today. I am a product marketing lead at datab bricks for SAP and I'm here to just get a lay of uh what where we're at with SAP and data bricks both. So, can I get a quick show of hands? Who here today is a data bricks customer or uses data bricks in some capacity? Most of you. Okay. Who here uses SAP today? Amazing. And then who here has tried to get SAP data closer to data bricks or to do some kind of AI ML analytics?
Was that fun? No show of hands. Right. There you go. We got this a lot from customers when we you know launched many new partnerships say hey what about my SAP data how do I get that in data bricks how do I do AI ML SQL analytics notebooks you know a lot of blending of SAP and nonSAP data and that's when we announced this partnership earlier this year and we're here to invite SAP to tell you a little more so Har welcome to uh the talk and I would love to welcome you on stage to tell you a little more about why this partnership is truly groundbreaking because we here for the first time are launching SAP data bricks data bricks' OEM product inside of SAP business data cloud and this is where the magic happens and you're here to tell me more about that welcome so this is going to be a different way of presenting because you don't hear the sound I'm expecting you all hear me.
This is super weird, but hey, we'll get through it. Um, so I handle all the BDC ecosystem integrations and product management. So today I'll walk you through essentially the SAP strategy overall that we announced earlier in the year. Then how data bricks and BDC fit in and then we'll go into a bit more detail on what exactly we do with data bricks. Yeah. So let me start with a basic scenario, right? Most most companies you're looking for this this northstar closed loop immediate decision making but the practical aspect of that is today you get data from your applications you load it into a data warehouse then somebody works on a dashboard gets some insights and then takes that insights then thinks about it and then he does something which results in further data getting back into the application.
2:15 SAP strategy and vision
Now this is neither super close loop or neither nor immediate right it takes time for an insight to get an action done and data getting generated in the application but now with the nexus of data and AI coming together much closer and with the access to applications that SAP has SAP's in a great position to really bring that together that close decision- making and that's precisely the strategy we announced fundamentally in a simple terms what the strategy Y is you know SAP applications generate enterprise critical mission critical data that data for the first time is being extracted curated and provided in a managed service environment called the business data cloud and this is huge it's the first time it's being done from all the line of business applications and that data is now made available for analytics and AI analytics is pretty clear we can do inside dashboard etc.
But with AI when the data is available for AI there are two important things that happen. One is you can use AI for human decision making you know decision- making that is assisted by AI which is essentially your co-pilot like capability and we call it jewel. So that allows you to look at the data and make a decision. Then in order to operationalize that decision at scale we have the agentic support. Yeah. So now with the data available for analytics and AI and with the support for agentic you can then land the data when you operationalize those decisions they interact with the applications which generate further data which then is sucked back into the business data cloud.
So essentially the strategy is centered on data powered by AI and enriched by SAP's applications. So that is netnet what we have announced earlier in the year. Now let me take a moment and go through the details of the business data cloud. So what what we have here is three layers. So as I said for the first time ever SAP is extracting the data from all its line of business applications and we land it in a common common format on the bottom layer called data products and harmonized data models.
4:36 Business Data Cloud architecture
So what we're doing is we're essentially curating the data and landing in a data lake that is managed by SAP. And why is this important? So today in most organizations you as a customer is managing all the ETL and there are myriad ways to get at that data. You know there is replication there are API endpoints and all of that all that kind of goes away and for the first time ever you're able to get that data from SAP landed by SAP managed by SAP.
There are a few important things that come with that. First is once you land the data as data products, you now don't have to hit the applications over and over again to get at the data. Uh for example, if you think of an ERP system, multiple parties in an organization want to get the data from there and they all tend to hit the ERP system causing a load on the operational system. Many CIOS don't allow you to hit the system.
So for the first time ever that data is available as data product in a managed lake environment from which you can share. Yeah. and there the data products are available as part of a delta share mechanism in in in the lake. The second thing is that data will be available with semantics for you and that is again um a very important thing because when you were replicating data out of the SAP systems you were replicating raw data out yeah you didn't get the semantics you had to rebuild the semantics and fundamentally in a simple analogy what that means is when you replicate raw data it only tells you what it is like quantity is equal to 100 but when you you really want to know what it means is it 100 pallets is it 100 packets packages.
That's what you mean by the semantics. So SAP is able to offer all that information with the semantics available in the data product for you to do these um instead of rebuilding it. And the third thing is now the data products will be available for downstream consumption using a delta sharing approach. So instead of having to replicate the data and manage multiple copies downstream, you can now share the data. That data is managed and maintained consistently by SAP and you'll be able to and by consistency I mean for instance if you take a purchase order data product it'll have header line items and all that we will manage the consistency instead of you having to replicate five tables and having to manage the consistency downstream.
So that's again a super important aspect of data products and they harmonized in that lake layer by SAP. Now the layer above that you see a series of engines. These are the engines that'll consume the data product and they come with certain innovations that we've landed as part of BDC. So the first engine that you see is datasphere. Datasphere was an existing engine which allowed you to understand the semantics of SAP data and allowed you to build model the data with self self-service modeling and then use the analytics tools over over that.
So with datasphere what we now done is we have added a lakehouse support under datasphere which means customers as they load data into it now they have a cheaper way of storing the data in an object store in a delta lake environment. So any data products they create themselves or derive out of the standard data products SAP provides can land under the data lake under data sphere. So that's that's an innovation we've added there.
The second thing is you see analytics and planning there what we have done is especially when it comes to planning today the challenge with planning tools is you have the actual data sitting somewhere and planning data sitting somewhere you end up copying and moving this data now for the first time the actual data and the planning data models can both sit in data sphere and the sec planning can work against that and that's huge because typically you end up moving the plan data now you don't have to you can you can merge both the plan data and the actual data together in one environment that is on the data sphere.
The third important thing is you see BW BW modernization it says so BW is our onrem data warehouse in the market for 15 20 years and it has significant amount of data from our ERP systems within that so what we're enabling is to move that into into a private cloud environment as part of BDC it'll become a trusted entity in BDC and make the data available for both analytics and a IML yeah So we'll we'll have an extractor within that where it'll be able to extract the data land it in an object store make it available for analytics and AI and that is where data bricks also comes into the picture in that till now with the engines we had available we could do analytics really well but now with data bricks combined into the mix we can do analytics and a IML together really well and especially with the 15 20 years of historical data that's sitting in the BW now made available in an object store in in BDC that becomes significantly useful not only for analytics through data sphere but for a IML with data bricks and data bricks of course there's a few other things we can get get to that and the AI foundation part is really enabling models that you create to be so we have a AI hub where you can um you can access all the different LLM SAP manages all that so that's a focus of the AI foundation where if you use data bricks to build a model in order to serve it you can deploy it to the Gen AI hub that SAP manages.
Yeah. So that's the middle part where the engines are there with innovations in that they will be able to access the standard data products in the layer below and you can create derived data products and with the with the addition of data bricks in the middle. The huge benefit you get is that for the first time ever a data analyst, a data scientist and a data engineer all three are able to collaborate. Previously if they had to collaborate you had to copy and move the data over to another data lake and they work there.
Now they can collaborate seamlessly in one single environment and that is kind of really an important aspect. Then the third thing SAP has done is we we will ship pre-built applications that are managed by SAP updated by SAP the configuration life cycle all managed we call them intelligent apps. So for each of the domains whether it's ERP whether it's HR whether it's spend for each of these domains we will have a series of applications that are completely managed by SAP in a managed service of environment and ship and we'll have a series of partners working with us who will also create these applications and ship.
So for partners the big thing is that there's a whole wide space of applications that they can create on top of the data products that are available that they can now build and ship to our customers. So again um an important area for both our partners, data bricks partners together to work jointly to build these applications on BDC because now you have data from every line of business application available in that. And what that what that really expands is the ability for you to do these cross analytic use cases that were not possible before because you had different ways to get the data from these backend applications.
So looking a little bit more deeper at um data bricks in the box in the BDC box what does it do there are three important things one is that you get the notebook environment completely you can work with that just like you you're familiar doing with that in data bricks environment second thing is I talked about that BW data and the data products right that large amounts of 20 30 terabytes sometimes data is available you now with data bricks because it can be those data can be shared with data bricks you certainly have the ability to do ad hoc SQL analytics against that data with the engine that is available and of course you can use mosaic AI to work on that data too and all the new data products that are being built definitely unity catalog is available so that you can have for downstream consumption you can manage and govern with that and everything is delta share in the box all the data products we land are available for sharing as a delta share and that seamlessly works with data bricks in the box so there is no data copying and managing It is shared seamlessly with the datab bricks environment in the box.
11:48 Databricks integration benefits
And what we'll also have is a seamless sharing of that data with the data bricks brownfield environment because many of you raised your hands you said you had data bricks in place. So we'll have that ability to share the data even with the brownfield environment. So whether you are green field that is if you want to use the data bricks in the box or you want to share data both both facilities will be available. Yeah.
So I'll I'll skip the slide because essentially one final thing on the on the customer is at a macro level what this does for our customers today if you think about it when a customer does such environments whether it is um a data bricks environment or SAP environment standalone and they're trying to build insights about 60 to 65% of the time is spent in that ETL getting the data prepared and you are getting to about 15% of the time spent on insights and this these are stats I've heard from customers and that's how I'm sharing it with the 60 and the 15% number now with the data products being curated and managed by SAP that number can suddenly come down to closer to 20 to 10 and you can now increase the amount of time you spend on the insights because the whole task of managing ETL suddenly kind of goes away and the analogy I would use is in in many companies if you think of the data engineers They sometimes live in shame and fear because the pipelines break.
13:11 Impact on data engineering
They're never sure whe the data is correct. They're never sure whether the report at the end is right or not. Now for suddenly with the data products being managed by SAP at least for the SAP side of the business, they can live in pride and joy because now someone is managing and managing it consistently for them. So that's the shift and that is a huge benefit for all the people who have to do the ETL from the SAP data which was anyway hard to hard to manage initially.
So what did we do with the with the integration of data bricks in the box? So with there is um when you get BDC there'll be a button in there say click here and that that onboards data bricks into the environment because you have to have data bricks as a trusted entity in our environment so that they can share data easily. So that's the first step. The second step is now you can we have a catalog where all the standard data products and derived data products are listed.
14:38 Integrated workflow experience
You can go into the catalog and right click on a data product and say share with data bricks. It is that seamless an experience and it would then appear in the unity catalog. Then once you're on the unity side and data brick side, you blend it with other data. You now have a capability there to create a data product to share back with SAP. That is the that is the third step you see at the bottom where SAP data products you can create the data product unity catalog and create a share which is then shared back with um with uh with um our business data catalog and now you can blend that data in the modeling environment that data sphere provides and you can run the dashboards.
So how where does this make sense? So essentially when you're building for instance I've got this example you see the treasury dashboard and the postings outlier. So this was an example where at the bottom all this ETL was going on. Much of that ETL now gets replaced by BDC. But previously in order to do the postings outlier analysis you had to copy the data to a separate data bricks environment do the analysis copy the results back and get to here.
15:51 Practical automation example
But now for the first time ever you can share that postings with data. You can do the postings postings outlier analysis and data bricks share the data results back and have it blend into the dashboards and show it up in the treasury dashboard. So that's that is a huge difference in terms of automation that you can achieve that was super difficult before. Yeah with that let me finish with the last slide. There are two two very important concepts that you as customers and partners want to keep in mind.
16:35 Data product strategy
The first is SAP will have a series of SAP manage out of the box standard data products from all its line of business applications. That is a step change in the way SAP has come to the table. But you'll also get the capability with the tooling we provide from custom managed from the if you have on-prem systems today or you have not gone onto our cloud environment you will still be able to get the data replicated into BDC create a data product out of that and then share it downstream.
So both these cases are useful especially for partners who are working with both of us the custom managed data products become super important because now because they have built a lot of IP with domain knowledge into our ERP systems that data can be extracted using the custom managed route brought that into BDC and then shared with both the data bricks environment and of course the analytical environment that's available within data sphere within um BDC so that really is where Both data bricks and the SAP technologies come together with the semantics now available to you as customers because the semantics are wrapped up in our data products.
Our data products are defined by a protocol called ORD and it is open source. ORD has three important things. It will it will have the metadata around the data set. It will define the relationships among the tables the primary key foreign key all those relationships are designed and it will have an endpoint. The endpoint today is going to be a delta share endpoint. So thus you don't need to copy the data you can share the data and share it seamlessly with the unity catalog.
So that really is the big picture story what we have as I said at SAP the whole strategy now is centered on data powered by AI and enriched by applications and now in this mix it used to be analytics predominantly with the addition of data bricks now it is analytics plus a IML workloads together for a customer in a in a nicely integrated manner. So that's that's all I had. I'm happy to take any questions if you have.
Learn more about the Databricks Data and AI platform.
The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.