May 3, 2024

Redefining Cloud Trust- Perspective of Schwarz Group

The Schwarz group is the largest Retailer in Europe. In the digital age, data protection has become a paramount concern for the Schwarz group and businesses operating within the European Union (EU).The complex and ever evolving internal systems hold vast amounts of data urequiring advanced and dependable infrastructure. This session will explore the challenges the Schwarz Group faces from their perspective and how they’re working with Dremio to build a secure, independent data cloud solution.

Sign up to watch all Subsurface 2024 sessions


Note: This transcript was created using speech recognition software. While it has been reviewed by human transcribers, it may contain errors.

Aman Kalra:

Good afternoon, New York, I’m Aman Kalra and I’m the product manager here at StackIt for the Data and AI platform. I’m hoping you were part of the keynote yesterday, where Benjamin, along with the CEO of Dremio, launched the new Data and AI platform. It’s thrilling. The response we received is overwhelming, and we’re excited how our customers will seek value from it. Today, I’m here along with Daniel, who’s a product owner at StackIt, and we would like to talk about the Schwarz Group perspective in understanding cloud trust and what are key actions in that direction. 

The Schwarz Group

At first, let’s have an overview of what Schwarz Group is all about. With Coughlin & Little in the forefront, the Schwarz Group is the leading retailer in Germany and boosts an impressive annual sales of roughly 154 billion euros annually, reflecting its significant impact on the retail industry here in Germany and its ability to meet customer needs across at a massive scale. We have a global presence in around 32 countries with over 1,700 locations worldwide, including supermarkets, hypermarkets, and discount stores, often well-known under the brands of Little and Coughlin. Little, I think, also has stores across USA, and we hope you love shopping from there. With over 500,000 dedicated employees that we have at Schwarz Group, we have showcased diversity and also how spread our teams and infrastructures are. A fun fact here, we have over 100,000 machine learning models that are running and powering our systems here at Schwarz Group. They help us forecast demand and fulfill them accordingly so that our customers go happy. 

Data and AI Workload Growing Exponentially

With the scale of operations that we just spoke about, come impressive capabilities to handle the technical advancement here that we face every day. Today, there are several petabytes of data that is generated on a daily basis, which further is used to analyze business processes, improve working capabilities, or discover new insights. Several models are trained to expedite the intelligence on and off the store, which brings in further capabilities. We have specialized deep integration of artificial intelligence and machine learning in forecasting demands and rightfully stocking our warehouses much in advance. Image recognition also plays a strong role, especially in our Schwarz production unit, where warehousing has a strong focus on the quality and the speed of products that we build. With the models at play, there’s a strong correlation of our business and other venue. We have several dependencies, our infrastructures that power the data centers, and capabilities of training these models. While these data centers and infrastructures are becoming complex, they’re also well-separated or segregated across different regions in EU. 

It should not come as a surprise to you, the scale of operations that we are doing here today. Efficiency is in our roots, and the way to be efficient is to leverage data on a day-to-day basis. For us at Schwarz, the value of data is ever-increasing. There is easy access or self-service capabilities where every department is now enabled to access data and have insights in no time. However, while we enable these services, the goal for data protection increases tenfold. As you know, Germany is part of EU and have strong relations to data and privacy. We at Schwarz are bound to adhere to them, adopt them, and respect them in our everyday businesses. Let’s understand the status quo of data protection at EU to get a fair understanding in a complex world that we live in. 

Protection Laws

Prior to 2000, it was Wild West. The dot-com bubble was closing and connecting countries closer, which opened doors for smoother data transferability. Then came 2000, where Safe Harbor was introduced. Safe Harbor basically was an agreement between USA and EU that simplified data sharing between these two places. It made it easier for companies to share and process data. However, in 2015 came in the shrimp’s one. This was a court case which was in front of the European Court of Justice that invalidated the Safe Harbor. It basically pointed out that the Petriot Act enables USA government to access any data on its borders. A year later, we came across Privacy Shield, which was introduced to replace the Safe Harbor. However, the court problem still existed. Well, as you know, it did not last long, and shrimp’s two came in and invalidated Privacy Shield in 2020. To tackle shrimp’s two came in the new standard contractual clause, which consisted of something called transfer impact assessment. The idea of transfer impact assessment was simple, that companies have to assess the challenges or the impact of transferring data into a non-EU company or a non-EU country. Sounds easy at first, but it’s a massive amount of work. You have to start considering all the regulations and legislations of the non-EU country where you’re transmitting this data. 

At the same time, even GDPR was launched, which is General Data Protection legislation. The idea of GDPR was to give a way which could govern how we use, process, or store any personal data. While it brings a clear set of rules on how data is handled, it still introduced a lot of complexity for the organization, which started to increase the amount of time and effort, and in the end, the money that they had to pay to maintain and adopt GDPR. While it’s still adopted to date, companies struggle and often need an expert advice to take care of their customer data. With Privacy Shield in two, which was introduced in 2023, it addressed a lot of concern that was raised with shrimp’s two. It restricted how US can or intelligence can gather certain information, but it was still a way ahead. We as a company are aware that there’s a lot of back and forth establishing certain safe protocols that EU establishes, whereas shrimps on the other side comes and invalidates it. We might just wait for a bit, and we introduced with shrimp’s three, which again, invalidates Privacy Shield two. So we at Stackit are able to understand this and have some thoughts on how we might approach it. 

Let’s quickly take a detour and dive what we have just learned into perspective. Let’s see how US, EU really takes their stance when it’s concerning data protection or in general, even implementing the plan it lays out for EU. Let’s consider the example of Apple. Apple had to replace all their lightning ports with a USB-C charging port, which I believe was not cheap. Now considering how Apple really focuses on large scale production, and now they’re trying to adopt this to entire portfolio of Apple. Cloud usage, in a survey found out that cloud usage in Europe is quite low when you compare it to America. We at Stackit see a lot of German companies wanting to move their workloads and their data to the cloud. Oftentimes it’s the slow adoption because of these tough regulatory requirements. Sustainable data protection is a continuous effort for us at Schwarz to maintain our infrastructure and keep progressing towards a sustainable journey. We have three simple requirements, stability, a stable ecosystem that enables continuous delivery and smooth operations of our infrastructure. Second is reliability. We need to have reliable processing, data processing capabilities that are within the legal framework. And the last one is our focus, that we always are focusing on delivering value to customer while we stay within the boundary of legal framework and the data security. Now I have Daniel with me who dives deeper into our efforts and mainly translating to meet these requirements that I have.

STACKIT: The German Business Cloud

Danyel Gros:

Thank you very much, Aman. To satisfy exactly those requirements, the Schwarz group took a new path, and that path is called Stackit. Stackit is the cloud service provider for and by the Schwarz group. And now let’s fill that name, Stackit, with a bit of life. Stackit started its operation around the year 2019, and since then Stackit is flourishing. The goal is to have 70% of all the workload that is running within the Schwarz group to be running on Stackit. And that by the year of 2027. So we have three years to go and we are right on track, but there is still a lot of work to do to achieve that one. That means Stackit is the IT service provider for the Schwarz group, but that is not the only goal or the main goal of Stackit. The goal is to be the cloud provider accessible for the whole of Europe. And we are working on that one every day. Now let’s deep dive a little bit into the requirements the Schwarz group has on its cloud service provider. There are a lot of those, but let’s talk just about a few of them. 


Massive resource requirements. The Schwarz group needs massive amounts of resources, storage, compute, network, whatever the Schwarz group needs it. On the other side, the Schwarz group requires its services to be simple, to be easily accessible to every employee within the Schwarz group. And Stackit has to provide exactly that. The Schwarz group wants to rely on major and stable services, which of course is a given when so much money depends on the operations on exactly that cloud infrastructure, it has to be really rock solid and Stackit provides that. On the other hand, Stackit needs to be quick when adapting to changes, especially when it comes to technological advancements, Stackit has to keep up to be sure to be a future proof. The Schwarz group has legacy systems. Those need to be supported by Stackit and the Schwarz group has lots of them. They are running well, but they will be migrated or are currently migrating to Stackit. So Stackit has to support them. On the other hand, we have the most modern cloud native applications that not only need to be supported by Stackit, but Stackit has to encourage the development of them. That is one of the main focus points we have, and it is really, really important for Stackit. So those are just some few requirements. To summarize them, the Schwarz group needs everything. Everything a cloud provider could provide, the Schwarz group wants to have, and that is a really great thing. Why? Because everything the Schwarz group requires is made available to external customers too. And that is the foundation Stackit is building on. 

Past and Future of STACKIT

Now let’s have a look at the last years of Stackit. What have we achieved? The last years, we built a reliable infrastructure to build further services on. We’ve covered the base. We’ve built data centers in and around Germany. We built infrastructure as a service components, supplying VMs, Kubernetes as a service, etc. to our customers. We have the most important core components. Those include identity solutions, network infrastructure, platform as a service components, and many, many more. Just the things that are required for every cloud service provider out there, and they are working now, and they are working great. And what is coming now is the thing that is more interesting to us here in this audience. It’s the data and AI stuff. The Schwarz group has many petabytes of data being scattered around, and those data source, those vast amounts of information are currently being migrated to Stackit together with the corresponding workload. There is a lot of productive data and AI workload being running on the Schwarz group, and it is too currently being migrated to Stackit. That includes our hundreds, thousands of machine learning models used for different purposes, large-scale analytics, workloads, and many, many more. Those are being migrated to Stackit currently. And one of the tools Stackit is providing to make that migration process easier is the Stackit data and AI platform. I can be one of the product owners working for the platform, and our goal is to provide exactly those managed services for our customers that enables them to offload their data and AI workload onto the cloud, onto Stackit. 

Now, let’s have a little bit of a deeper look into the Stackit data and AI platform. At its core, we have a Lakehouse architecture with a new edit space. So that is really the core of our data and AI platform, our Lakehouse. And attached to that, we have different services providing different kinds of functionalities like ingestion, processing, visualization, and some more. When it comes to ingestion, we are especially focusing on having an easy-to-use application and where it is easy to attach especially new data providers to, since that is what enables the organizational scaling for us. And besides that, the support for large-scale workloads must be there, obviously, and we are covering that as well. When it comes to processing, it needs to be really easy to spin up new clusters for some large-scale data crunching of some sorts, and we are doing exactly that. The same for visualization. It’s getting especially important when we’re talking about data analytics, AI analytics and stuff, and it’s really important to have some visualization options available here, and Stackit, again, is providing exactly that. 

Achieving Our Vision

How do we achieve that? Well, there are different focus points we have here. One is we are quickly developing our solutions together with our friendly customers. We have an extensive pre-sales and consulting effort going on here to build reference cases together with our customers. Our goal here is to understand the problems, understand the requirements our customers have, and build solutions out of those. With that knowledge we gain here, we mirror that one into our product development efforts where we can use that knowledge to build exactly those products and manage services that our customers need and that will fulfill their requirements to be just a great product to use. And when we have that product available, we can use it together with the scaling capabilities of our partners. For example, we have partners that we are working with who, in turn, are working with multiple cloud providers like Stackit, which allows them to build the best solutions for their customers, which in turn allows us to scale Stackit into new and different areas. Everything we are talking about here today is not just some future vision, but it’s really happening right now and it’s really amazing to see how fast Stackit is progressing here. I’ve brought a few examples of things that are happening right now. We have our partner Abinitio. Abinitio provides a high-volume ETL/ELT platform for enterprise use cases and the Schwartz Group has been working extensively with Abinitio for the last years. And now Abinitio is already running on Stackit. And I think that is really, really amazing. We have an internal project, the so-called Unified Analytics Platform Cloud. It’s an internal management platform to be used by our internal employees to provision projects, tools, environments, networks, and stuff for data and AI work. So it makes it easy for our internal employees to provision those tools and it takes away some of the management overhead that way from the employees to make sure they can start quick and easy. And this UAPC is currently migrating to Stackit and I’d really love to see it. We have the Schwartz Inventory Management and that one is the heart of the Schwartz Group. As you can imagine, it is one of the largest and most modern inventory management systems out there, just like it is required for a large-scale trading company the size of the Schwartz Group. It contains really massive amounts of data and has a high need for large analytical workloads. And again, it will migrate to Stackit. And I really like to see the progress Stackit is making here. It’s just amazing. 

The Bigger Picture

Okay, now let’s take a little step back and look again at the broader picture. Companies are currently facing different kinds of challenges. Most of them are known to you as well. We have increasing uncertainties in different areas. We have a market pressure that is ever increasing on companies, which leads to the need to more and more focus on what is important to bring value to their own customers. And for most companies, that does not include stuff like building infrastructure or worrying about data protection laws and processes, but that’s a key feature of Stackit is for. The most important thing we want to talk about today is that it’s especially important for European companies is sovereignty. We want to talk about the topic of sovereignty a bit more. Currently European companies are being pressured to decrease their dependence to the outside. Political changes in the last year have made that clear, especially with the strict data protection laws, it is getting harder and harder for European companies to interact with other legislations. And that includes American infrastructure and American companies as well. It is generally not known how the political landscape will change in the next five years, 10, 15, 25 years. But the general assumption among European companies is that they will be better off when partnering with and depending on European companies. And that is currently creating a gap within Europe, a gap that Stackit is filling. Stackit is providing exactly those cloud services that are required by those European companies. And our main emphasis here lies on especially one thing. We want to bridge that gap to Europe and to European companies together with you. 

This talk here is not intended to be a sales pitch. We want to put our main emphasis on one point, it’s our partner program. We don’t yet have all the answers to all the questions that are out there, and especially to the questions that are yet to come. But in this audience, we have so much knowledge and so much know-how gathered that we really would love to be working together with you and partner with you to bridge the gap to our European customers and to your potential European customers. We are currently working together and partnering together with Dreamio. And it’s really, really working out great for both of us. Whether you want to support us when we are building services for our customers, especially in the data and domain, or if you want to build solutions on Stackit for your customers or together with us for our customers, please contact us. You have the QR code here to the contact form. You can use it, you will be redirected to the correct team within a second. And I can only say again, please, if you want to scale out into the European market, we are here. We would love to be working together with you. Thank you very much.