Optimize SAP-based supply chains and improve sustainability with AI/ML

Many customers have been asking us how they migrate and modernize their SAP systems using AWS native services. Once on the cloud, they want to understand how they can extend their business processes using innovative services such as AI/ML to unlock value for their business. If you are one of those customers, you're in the right session.

My name is Joe Curri and I'm a Principal Solution Architect for SAP here at AWS. Today, I have the pleasure of co-presenting with the Erickson team who are going to take us through their migration and modernization story on AWS. We'll be hearing from Johan Sporer, VP of IT Cloud Services, and Rickard Väsfors, VP of Automation and AI.

In today's session, we're going to be looking at why customers are modernizing their SAP environments and doing so with AWS. We'll then turn over to Johan who's going to take us through the Erickson migration and modernization story. He'll take us through the steps his organization followed and also some of the learnings they had along the way.

Rickard will then take the stage and explain how he's led an initiative to embrace intelligent automation and AI on top of the foundation of SAP running on AWS. And he will show some impactful use cases around supply planning and transportation management processes.

And from there, I'll briefly touch on some of the AWS services that you can use to follow in the footsteps of innovators like Erickson. And then at the end, we'll have time for some Q&A.

So working with thousands of customers like yourself, we understand there's a wide array of internal and external challenges and you are seeking solutions to better your business. You may be driven by compelling events like cost objectives, data center exits, M&A activities or end of life for your hardware and software.

In addition to these urgent drivers, you may be strategizing on how to transform your business and remain competitive in the market. The conversation is changing. Customers expect more than just cost savings and efficiency. They want to move faster, make better data driven decisions and transform customer experience. They want to deliver digital transformation and modernize businesses. They want to reinvest cost savings to drive innovations and new business models.

And you will see some of these drivers today in the Erickson journey which Johan and Rickard will take us through. And taking advantage of the AWS cloud to run mission critical workloads like SAP has been an effective way to address these challenges and achieve these aims. It provides performance improvement, cost savings, improved reliability, security posture and agility.

So customers who address these challenges and accelerate their business transformation typically follow the following stages:

First, they run their SAP workloads on the cloud benefiting from that performance improvement, reliability, cost savings and agility. Customers then have the option of which path they choose to migrate. They can simplify their journey to S/4HANA by adopting an integrated solution like RISE with SAP. Or they can migrate an existing SAP system from their on-premise data center. Either way, AWS has been running SAP workloads in the cloud for over 15+ years. And customers are able to leverage our extensive APN partner network, training, purpose built tools and even funding. It's all designed to help you make the shift in a way that minimizes disruption and gets you back to managing your business, not your migration.

Once in the cloud, customers have the opportunity to unlock the value of their data and deliver new business insights. Often this begins by adopting an approach of combining SAP and non-SAP data. And this gives you a true 360 degree view of your customer and business processes. Once unified, democratizing access to analytics and delivering insights to users where they can make better decisions is possible. And today we have over 200,000 data lakes running on AWS.

And lastly, customers who follow these steps can innovate faster and transform their business. Customers have a real opportunity to use the added agility and innovative services from AWS and SAP to transform their business. They can extend processes into other systems, create custom applications and optimize core business processes all while keeping the core clean. You'll see shortly how using some of these 200+ services including cutting edge technologies like AI/ML can really have a transformative effect on your business.

We've been helping thousands of customers on this transformation journey across numerous industries. We're fortunate enough today to have to hear from leaders from one of those customers, Erickson, and their transformation journey really embodies what it means to migrate and modernize an SAP landscape. I'll play a brief video before welcoming Johan on stage.

[Video transcript]

Often when I talk with peers in other companies, we discuss the speed of technology. It's an ever increasing speed of technology that we're facing as technology leaders. And one of the important parts that we play as technology leaders are to demystify new technology.

Sometimes we face technology that is known like cloud, AI/ML, accelerators and sometimes the technology that we are facing is more novel, creating a lack of understanding of the structures, lack of talent, lack of how to implement and how to actually scale these types of technologies.

And I think it's as leaders, it is our job to demystify the technology, finding the one or two or three applications that we can rally our employees around and really do proof of concept and then scale and then implement as a run into the business and really create value and unlock some of these technology promises.

So our view is that the speed of technology is ever increasing and also the benefit and opportunity of technology is ever increasing. It is crucial that we provide this guidance to our employees on how to use the technology and how to adopt new technologies.

The question that we find often is how to really leverage the community of employees and really get them to embrace and adopt the technology. I'm a firm believer that the companies that actually enable our employees to adopt technologies the fastest are the winners in the market.

So the transformation is not about technology, it's really a people centric transformation that we're facing.

So our vision for the IT transformation was we really needed to exit our data centers. We needed to change our entire ways of working to enable the cloud benefits that we could see from implementing the new technology.

We really needed to prioritize our transformation and we really needed to harness the power of automation. It's about setting clear and ambitious goals, quantifying objectives that really reflect our commitment to move into these technologies that set the transformation apart at Erickson.

We set a brave target at the time, we said 80% of our compute would move to cloud and with setting that type of high target and ambition, we tipped the scale of which culture would actually prevail.

So we had to transform our culture, learning the new ways of working. Then we embraced the product centric agile and DevOps mindset in the organization that enabled us to adopt the technology much faster in our adoption strategy.

Our organization change was adopted an incremental change. So we view this through the lens of a product and product portfolio allowing us to be more flexible and responsive in operations.

So we could actually move and build the platforms with a product mindset doing changes stepwise. So that enables us to really launch innovations, try and do errors and then quickly recover.

So as we look into our aim, we aimed at moving 80%. We are now at 90%. When we set the target, people told me it's impossible to move 80%. This is four years ago. They said 50% is an ambitious target.

I think I was told as well, you couldn't do lift and shift, lift and shift would be expensive. It wouldn't work. We did a lift and shift and we actually cut the cost with 20% below the business case that we looked at. So it's doable in our transformation, we really needed to set up a factory that worked.

So we established a big factory that actually focused on enabling in the product organizations to really take a decision within the factory. Often we encounter outdated processes, directives or policies that are blocking the technical transformation.

So the advice is to really try and keep the decisions within the factory. As you migrate, the duration of the program was 18 months from covering detailed solution design, technical foundation and then migration to AWS.

The time we had the time constraint of leaving the data center which really forced us to do a lift and shift. We then migrated about 50 applications in SAP, about 600 AWS instances was deployed.

The migration was divided into five subdivided waves grouped per application area technology systems - SAP ERP, analytics, sourcing, CRM. Within each of the waves we actually learned because we moved first development, then QA and then production.

This way we actually adopted a way of learning as we moved and we also focused a lot on automation. So as we moved from development to quality, we could actually adjust our automation skills, we could trim the way we actually deploy the solutions. And as such, we minimize the impact of business.

The incumbent vendor and operational service provider was responsible for testing and integration. We had weekly status calls that gathered all involved parties where we actually focused on blockers and releasing some of these processes or directives that could actually be cumbersome for the factory to manage.

This is the architecture for SAP applications on databases. It's a multi availability zone and a single read architecture. When we consider the future design, we had some firm principles that we followed:

• We followed AWS architecture best practices.
• We deployed SAP systems focusing on high availability and reliability.
• We had established a clear design on clustering and redundancy, implementing cluster and redundancy for SAP components such as SAPAC, SES and NetWeaver and HANA databases to limit single points of failures.
• We utilized AWS Availability Zones to achieve high availability and SAP systems distributed across three multiple availability zones that actually then physically segregated the data centers with the redundant infrastructure.
• We used Lemongrass as a cloud native partner to leverage their automation capabilities to actually move this migration much faster.

Benefits and migrations:

So what happened was after the migration, we had a program meeting and today's program meeting, one of the financial direct reports to the CFO came down and he said next week when you migrate the entire systems, I want regular updates. And for us that was the best grade we could actually have because he actually missed it.

We actually migrated the system the weekend before. So it was super seamless and I had that cup of coffee. I will remember for the entire life. The migration was super complex. We had 300 people working 24 by 7 across the globe in different areas. They were booked into different hotels and it was a minute by minute planning.

So we had people dialing in 5 minutes before they were due to do a change and then confirming that they were up and running. And at the end, we knew that if this system that we move this weekend would fail, it would be a top line impact and it would impact the stock. So the pressure was super high for the team.

So now we will hand over to Rickard to take us through the transformation powered by automation and AI.

I'm Rickard Väsfors and I'm heading up Enterprise Automation and AI. And my purpose and the purpose of my team is to radically transform Erickson through intelligent automation and AI. And what I will focus on is how do we use this great technology Johan and his team has put in place in reality.

First a bit of background to what we're doing. So we have for many years invested in exponential technologies like machine learning, blockchain, low code, RPA and combining that with an AI transformation approach. And the reason why we're doing that is that we see that we can get the factor 10 improvement in customer outcomes and business outcomes and customer experience.

We are also using this technology to learn from internal and then develop products and services that we can use to sell. However, this is quite difficult. And we did a survey 2.5 years ago where we asked 2500 decision makers on how they are scaling intelligent automation and AI. And 99% said that they had problems getting away from the POC stage and they only weren't able to scale.

I don't know if that's still relevant, but when I meet companies, I see a lot of POCs when it is about the machine learning and AI and a lot of talk as well. And one of the key factors here is of course, that they don't have a scalable platform.

The way I see it, we worked with this since 2016, we have 300 solutions live across the whole company basically. But still, I only think we are in the beginning of this exponential journey. But we have still got some recognition, for example, just a couple of months ago, Gartner named us as one of the leaders in what they call the Gartner Hyper Automation 100 Club.

And that is the companies who has the proven highest value when it comes from uh intelligent automation and a i or hyper automation as they call it also world economic forum recognized as leaders when it comes to uh industry 4.0. Uh and uh we have got something called a global lighthouse for our uh five g factory in lewisville, outside dallas, where we used a lot of our uh intel and uh in intelligent automation and a i uh capabilities.

So then if we start to uh how have we been able to scale 300 solutions? And one of the key factors is to have a scalable platform. And here, the partnership with a a double s has been great. Uh so what we did, we put together an um uh a iml platform, we call it amp. Uh it's using uh sage maker uh technology as a basis and also uh lambda and some other things. And this could be mainly explained through three major themes you can say, so you have enablers at the highest level and that's very much about making sure that you can scale uh solutions.

The second one is around operation excellence that is about maintaining the solutions and make sure that you also link it into our api structure and the digital infrastructure of erickson. And the last theme is security by design. We have for many years worked with responsible a i and we have a very strong framework for responsible a i. And what the amp does, it sort of forces our developers to follow our responsible a i practices by having built in responsible a i in the different phases of the development cycle and a quick look about uh in the infrastructure uh or the architecture and the architecture is based of the four different phases we have. When we develop machine learning, it's about link to experimentation, then it comes training validation and deployment uh deployment. And one key thing about this is that this uh architecture enables the fusion team where you work the way we are working. So both the platform team can look at the models as well as the different uh uh business people. And that is a very strong uh feature of this uh uh a iml platform.

So uh some examples of use cases we have delivered. And as I said before, we have done 300 solutions and they are basically addressing customer experience and customer experience. What that is about is very much to let our sales people to be able to put time with our customers instead of a lot of administration employee experience, we want our employees to have the same experience uh at erickson as they have, for example, if they do shopping on amazon.com, for example, the third area we are addressing is responsible business. And in that we have two major things, it's uh sustainability and its uh compliance. And there is one case which i mention here, which is about transport planning, which i will deep dive a bit on. And then the last thing we are addressing by our solutions that is efficiencies and that's basically cost out. And if you look to intell and automation and a i there is a huge opportunities in all companies for doing uh cost out and efficiency.

So let me start with one case and then i go over to the transport planning case. And here the problem we had was matching basically the demand and supply. Uh we had difficult to get visibility, what uh we needed to produce in our, in our, in our factories and uh for the teams to have uh visibility and to be able to have a uh a data driven decision making. So we put the place in a solution that uh utilizes both machine learning but also uh low code. Uh and it's now used for 82% of the ericson hardware portfolio. And it had se has several impacts uh on um uh for example, the uh scrap levels, inventory levels, et cetera.

So what value has this created? Uh so it's increased the forecast rail but with about 50% which is quite amazing for the last 29 months. The value of reduced inventory levels, reduced scrap levels and reduced um uh transport uh planning or, or uh utilizing the right transport facilities has created a value of around $50 million. And the increase of transparency during the whole supply chain has been also fundamental and really changed the culture and the way we are working. So that is one great uh uh case where we're using the a iml platform.

And the second case, it's linking them to transport planning. And what is the problem here? The problem here that we had difficulty to have visibility more than five days for ordering different transports. As you know, we are working in 180 countries. There is a lot of transportation of our products. Uh by putting this, this is a machine learning solution in place. We have increased the, the visibility uh uh quite a lot and the effects of that makes it easier. We do sort of rfps every second week on transportation if we're going to have a flight uh freights or, or, or ship shipping by boats, et cetera. And by doing that, we achieve um much better choice of transportation. That means that we don't need to use a plane when we can use a boat and that has given savings of $4 million. But even more importantly, what this is is uh uh addressing is our targets towards net zero. Of course, there is a huge impact on co two emissions if we are using the right uh transportation and also improved uh leave times. So that was a quick summary of how we have used the cloud to build uh an a iml platform and some examples of uh use cases in this area.

So please, i went over to you again to summarize. Thank you. So, um what did we take away as learnings then in the sap migration and in that remit? So they focus that on the learnings that we have. And the focus that i would recommend people to have is the selection of partner. When you go from a traditional um it infrastructure, you, you have a certain set of partners. But we recommend that you think about the capabilities that the partner have. And we would recommend to say if you migrating your sap workloads to cloud, the partner should have a demonstrated capability and knowledge in both domains in the same person. So the people working on it should know aws cloud and sap at the same time, then you can leverage the most.

I think also when you look at how these migrations can be done, you need to look at what assets you need to invest in and to get the stable migration and also stable operations. In the end automation is the key word. You know, if you can go to zero manual deployment, i think that should be your target. That gives you the ability to when you move from test or quality to production, you can develop your scripts and you can tweak and you can reduce the exposure of the change and the change windows of the business. It's a win win.

Also as you migrate and you migrate sap to the cloud, you're not the first company, right? So this is not the type of case that we talked about where the structure is unknown or skills is scarce or you know, this is where there's actually a best practice and my advice is follow best practice. It's been tested thousands of times and every company says we're so special, but i bet you're not really in this sense. So trust the team, trust the plan. And i think that's what's um will make you successful in your uh sap migration to cloud. Thank you.

Thanks johan and rickard. That's a really compelling story about how to transform your sap landscape. And once once they use the the benefits of the cloud on top, so before i um before we look at the capabilities, aws provides to innovate, we'll look at some of the requirements, the common themes we see at customers who've been through a successful transformation journey.

The first been reimagining their business processes. Customers want to move away from brittle systems burdened with legacy customization. They want to adopt industry standard best practice and they want to do this by adopting a clean core approach to their er p transformation.

Secondly, they want to unlock new insights with data and analytics. Customers want to break down data silos, provide users with insight, the insights, they need to make better, more informed decisions and they want to employ powerful a iml models to provide predictive and prescriptive actions.

And thirdly, customers want to develop a culture of innovation. They want to empower their creators with purpose built tools. They want to enable their teams to extend their sap system in innovative ways without risking the core. And customers address this challenge by adopting a clean core approach.

This is our vision for the sap clean core cove with sap. So rather than customizing at the application l level, customers extend their business systems at a higher level by integrating into the 200 plus aws native services and the 80 plus sap bt p services which are run on aws. So this enables customers to innovate faster with more freedom and deliver value for their business.

So customers have access to the 200 plus native services on aws to fuel their their modernization journey. This is significantly more services and with more feature depth than any other cloud provider. From the infrastructure technologies like compute storage and database to the innovative and emerging technologies such as machine learning, artificial intelligence, data lakes and analytics. And this makes it faster, easier and more cost effective to innovate and extend your sap system.

We've seen from the erickson story, how they use services like sage maker and data lakes to deliver rich insights for the business. Now with the aws sdk for sapab a, you can tap directly into the innovation potential of these 200 plus services directly from your sap system. The sdk makes it easy for a bop developers to extend and transform their sap based based business processes by harnessing the power of aws native services directly from sapab a.

So that with the stk, you can accelerate process innovation by harnessing those 200 plus services. It reduces architectural complexity. So you don't have to manually configure complex integrations between sap and aws services and it's done in a secure way. So all payloads are encrypted and it enforces the sap permissions. Um also, and it also fosters a culture of innovation. So it ups skills your sap a app developers and lets them use and upscale in the 200 plus cloud services available.

Now, the sdk provides access to the entire set of aws services so you can access the same a iml capabilities that are helping over 100,000 customers today innovate within their businesses. And with these capabilities, you can follow in the footsteps of innovators like erickson and use a iml in conjunction with sap to reimagine your core business process and nowhere is the is the promise of a iml in business transformation more exciting than that of generative a i.

So this enables you aws recently launched bedrock and this is now available in the, with the uh aws sdk for sap a up. This enables you to bring the power of generative a i to your sap business processes using the sap ab up language. Amazon bedrock is a fully managed service with a range of powerful foundational models from companies like anthropic meta stability a i and amazon's own titan model and this enables customers to quickly experiment with these foundational models and integrate them into their applications with the familiar aws tools and services.

And we've already seen customers innovate with this in our regular about packs you may have seen on linkedin. And we also um have developed use cases internally around dev ops bt p development, um improvements to sales and otc processes and integrating generative b i into sap analytics.

So not only do we have the 200 plus native services which you can use, we also have um 80 plus sap bt p services um available on aws. And these are powerful innovation services from sap, which you can also use to adopt a clean core approach.

So we have more bt p services running on aws than any other cloud provider as you can see from the graphic. Um and we have 19 services which exclusively run on aws and they, they include some powerful services like uh build apps and a i core which run exclusively on aws. And we have sap, have developed 100% of the new bt p services in 2022 on aws um versus 33% for the next um cloud provider.

And you can also use private link which enables direct or encrypted access from the bt p services to your a select aws services. So to meet your security and compliance requirements. So no matter which way, which direction you take um aws as you covered with the, with these bt p services running exclusive or running on aws.

So in closing, we want to thank our speakers from erickson for sharing their story with us today. So reflecting on their story, they've embodied what it means to modernize their sap environment and transform their business. The learnings we have seen um are common to a lot of successful companies have been on a similar journey.

So they are setting a vision for your team for what it means to be a world class enterprise, establishing your sap landscape in the cloud to benefit from the scalability, performance, cost saving and agility and extensibility it provides and finally creating a culture of innovation and a foundation for ongoing development, leveraging the best in class cloud services to innovate faster and transform your business. And of course, to choose the cloud provider, which has helped deliver the same transformation journey for thousands of customers. over the 15 plus years, we've been supporting sap workloads in the cloud.

So i'll just touch on a few um a few more slides and then we'll have the q and a. So i'd like to just point you to um some of the workshops that we have um for um ryan's best practices, some discovery workshops and our team would love to meet with you and discuss the best way for you to modernize your business. So you can sign up for those workshops through the key r codes.

Also, i'd like to point to some of the other sessions which are happening this week on the sap track and highlight ent 225. So why industry leaders choose, oh, sorry, why industry leaders choose aws for ride with sap. That's on wednesday and join us on thursday for ent 229, which is using aws services rise with sap and um bt p to innovate faster.

And then we also like to point you to the what's new uh for aws for sap. So keep you can keep track of the new announcements which are coming out for sap. Um and a link to that page is provided and then finally be sure to visit us at the enterprise workloads kiosk at the aws village in the venetian. Um so you'll be able to speak to an expert there and learn how to modernize your enterprise workloads, not just for sap, but microsoft vmware and mainframe.

So with that, uh we'd like to thank you for attending the session and um we have marcus in the, in the crowd who can uh who will be moderating the q and a.