AWS re:Invent 2024 - Unlock your innovation potential with AWS, SAP, and generative AI (MAM222)

Perfect all right well we're going to get started I wanted to say thank you for everyone for coming today to learn how you can unlock Innovation with sa and AWS so we're going to walk through some use cases um but first what I want to talk about or just if you can show by sign of your hands is how many of you have heard of generative AI I'm just kidding right well how many of you have started to look at potential use cases all right that's good great and how many are you are might be working with Partners today on use cases okay all right a few that's really good all right anyone planning on barking on a POC prototype or move to production awesome nice anybody with production use cases today production use cases can you stop and give us some more insights when you uh would love to hear more yeah great now we're going to walk through a few things today so we're going to talk about our collaboration with sap on generative AI uh we're going to talk through some use cases and tools to help you get started on transforming with generative Ai and then we're going to show you some partner use cases that we're collaborating with partners and then how you can get started with some of the tools and help from our teams as well so by way of introduction my name is Beth sharp I am a a worldwide Tech lead for sap monitorization I've been at AWS for about two and a half years now but be prior to that I was actual sap customer implementing data analytics configuring for about 25 years and I was also a finance person so I love being able to enable business with technology and that's what we're going to try to do here today fantastic U my name is AA Calo I'm a customer solution manager at AWS and I work with sap in the Strategic ACC team I've been with Amazon for 10 years and my background is also in finance we were talking yesterday about that and how the generative AI is a great technology to to help customers innovate in in this space as well thank you El good thank you Beth team my name is solid Khan I've been with Amazon for almost six years and I'm responsible for our solution architecture delivery for sap workloads and I also lead our Global Technical alliance with that I'll get started uh thank you Beth for the intro and Olga for the introduction as well so we've got a lot in plan for you in this session we're going to cover a few topics like Beth mentioned we're going to go through our journey with sap and delivering Innovation around generative Ai and uh collaboration with delivering some joint reference architectures we're going to show you some real world examples that you can possibly play with today if you're an SCP customer and you're familiar with some SCP and AWS Services as well I'll go into some real world examples and I'll just before I go into the examples I want you to know that this entire reinvent reinvent week we're providing you an endtoend Journey for leveraging AWS services with sap you're going to be learning about uh new Services we've announced around assessment and Discovery in the pre sort of planning phases you're going to hear a lot about on migration assurance and migration support as it relates to actually deploying some of those workloads for sap onb and we'll also talk to you about some go live support as well so look on look on those sessions for your catalog and make sure to attend when it comes to generative AI it is surely a service that is inventing and Reinventing itself day by day not a day goes by where you don't hear about a new llm or an FM not to mention some great announcements that came out just yesterday with Andy jasse and Matt Garmin we learned about Project Nova and the family of FMS and llms that are available to everybody in Amazon Bedrock we'll touch on that in this session too two great examples of our customers parcel and Adidas leveraging generative AI to not only reduce waste in their inventory in their business process but Adidas a shoe maker that's leveraged Foundation models to train over 150,000 unique shoe designs to be able to inspire and to create brand new shoe designs for their Market this is an example that you see on screen where they're able to leverage um a famous Spanish artist named gudy to inspire brand new shoe designs in the area of generation and inspiration for sorana customers so let me talk to you about the collaboration when you think about end toin use cases we are honored to share for over 17 years we've been working with sap and Christian Klein announced even at Sapphire how important it is to partner with sap and AWS at all the levels for generative Ai and services building with AWS that's resulted in the release and announcement of our Flagship software that allows customers like yourself to be able to build bespoke generative AI applications with a service called Amazon Bedrock if you go today and if you're an SCP uh customer let me actually show by hands how many of you guys are leveraging SCP btp today in your environment that's that's really great that's about half of the audience here today you all have an option to go into btp generative AI Hub and select Amazon Bedrock right now and be able to build against two of our uh models in fact three now Amazon Titan which is our FM for image generation and summarization anthropic cloud and Nova to be able to iterate and build some of those use cases today within your sap environment and connect that through our connectivity with btp with AWS data as well we also have uh availability for AWS trainum and infinia chipsets that a sap is using to build their next Generation business AI application we actually have a proof point where SCP Engineers were able to leverage this chipset to train their models in less than two days over several days for their build on other chipsets and of course scp's jeel application is their generative AI agent that allows you to do a lot of functionality as it relates to abop generation as it relates to asking intuitive questions to answer around your data this service also runs across 5 to seven data centers on AWS so if you have a camera phone this is the time to pull up your phone and scan that QR code this data this uh diagram actually got updated literally yesterday and the latest blog is available for you to look at Nova models family of models including anthropic cloud and Amazon Titan we're super proud to share this joint reference architecture it's a guide that allows you to see how you can leverage our Amazon Bedrock service within a secure and reliable environment for sap business technology platform in building and iterating exploring your generative AI applications you will have in the future availability for agents and also knowledge bases as we continually iterate with our partnership with sap another key example here is if you wanted to go and explore this will guide you to the physical blog that discusses The Joint reference architecture with links to the Discovery Center missions how many are you familiar with Discovery Center oh I love it so if you go to Discovery Center you can look up a mission for AWS bedrock and in there it'll show you step by step how to activate the service within SCP btp we'll go into a little bit detail with Olga and Beth on the value of leveraging Amazon bedrock and Amazon Q I wanted to leave you with a note that Amazon Bedrock is a secure service when you're building against the FMS that are available within Amazon bedrock and btp your data is is not going back to the FM provider it's not going to go back to Titan it's not going to go back to anthropic cloud or to the Nova family it creates a private instance for you as only available to to you and your company as you make it available through your ACL and once you're done that dat can be done away with and nobody else will have access to it a couple of example on how we're able to produce uh some use cases for customers one in particular is for detecting and reducing downtime using AWS iot services with Amazon bedrock and SCP btb Services if you think about the the cost of maintaining and um interrupting of your business flow in your supply chain your factory you're able to reduce that time significantly to near realtime detection and early detection by looking at patterns of services patterns of behavior for your devices in your shop floor AWS I allows you to stream that data directly to your sap environment where you can build a module and we have this demonstration aail able for you also to detect anomalies within your shop flow process based on an anomaly you can create an action within your Enterprise asset management software to go with early detection with a notification to the plant manager to go take an action with either your belt with either your process or one of your engines for remediation you can apply this with confidence because the delivery of this architecture is available to you in a joint reference architecture as well now I'll touch on a couple of lines of businesses that sap has we'll touch on CX first which is an array of toolkits for you to imagine the customer experience by building not only um cool images but also new product details by leveraging customer insights around intelligent Discovery content automation are is anybody in the room leveraging CX in their environment today if you explore CX you'll see that they are also embedding the use of Amazon Bedrock to create new content be able to launch products faster and increased conversion rates from marketing all the way to uh consumer adoption with that I'm going to hand it over to Beth to talk to you about additional use cases on transforming your business Journey with generative AI Beth over to you thank you oh you're welcome great so thank you Sal lot um for walking through the use cases we're collaborating on with sap um we're doing some additional use cases outside the gener AI Hub to help our customers um use gener of AI to transform their business so when we talk to our customers about gener of AI over the last year year and a half um a lot of customers come to us and they want to know well how are other customers using it what are some use cases I know what generative AI can do I know the technical capabilities um but they want to see how to connect that to their business processes or priorities their challenges and their strategies so hearing that we worked backwards from common business personas as well as common processes across sap customers and worked backwards to connect them to T of AI capabilities to show them how to connect the technology to the processes that are common across sap customers what's great about that is in our workshops our Merion days and discussion with our customers they can easily see how generative AI can be applied and what's really exciting is they're taking it one step further and applying it maybe to a similar use case with a different set of data different set of process that will bring them differentiated business value for their business so it's just the way we've approached it and we're going to go through some of those use cases today to show you what's possible and hopefully give you some ideas as well so what we're going to first cover here is the technology behind this um You probably seen this slide if you've attended some of the Keynotes and some other sessions but this is how AWS approaches generative AI we want to make sure our customers have a full stack of capabilities so on that top level you have Amazon Q um that is where it's an application where you can use it where you're where you work for assistance um with generative AI assistant for your data in the middle level is bedrock as suat mentioned is where you can build and scale your generative AI applications with all the capabilities that come with bedrock allowing you to customize it for your use case and then at the bottom layer is an infrastructure um and as solat mentioned sap we're working with them and they're using trinium and infuential chips for their future business business applications or business AI so when you look at sap generative AI capabilities a lot of them are with bedrock and some with Q so with bedrock is where we're developing a lot of our use cases um what's great is tens of thousands of customers already using Bedrock for their generative a Transformations and capabilities you can see on here with the different logos it's across every Geo and many different Industries so we want to show you how to use this for your sap business as well with bedrock um one of the benefits as solat mentioned um by using bedrock and this is why it's used in a lot of our use cases it's easy to scale and build differentiated applications for your business your data um with that you have the choice of leading Foundation models which is great you can apply the right foundation model to your use case you can customize that model and you can bring your data to the model using rag um you can also use agents and there's a lot of announcements today and yesterday around agents allowing you to execute tasks based upon your data and processes and that is all done with security and privacy as soot mentioned earlier so one other uh way thing I want to introduce before we talk about use cases is our AWS SDK for sap ABAB and I believe we have a session on this as well um throughout the week but the SDK is an easy way for our customers to start accessing not just Bedrock but our over 200 plus AWS services and it's we seen a lot lot of the use cases is a combination of not just Bedrock but some other use cases that allow customers to truly innovate and transform and you're going to see that in the accounts payable um uh use case that I share in a little bit but it's an easy way for you to access and it's where customers really get started to experiment so as again I said we approached the use cases that we started developing internally with our customers with common personas and common uh business processes to make it real so people can start to see what's possible so when we started looking at these are just a few that we started with and using generative AI for sap and nonsp data and processes you can innovate you can increase production and productivity for your internal users as well as your Tech users um and gain insights from those so a few examples where you can start with would be executive insights um we have people looking at uncovering critical insights from your SAP systems and your sap documents by using bedrock to summarize that and then sharing it out with your team saving time and making data driven decisions Finance managers you know Finance people are always having to produce the reports that really run your business and a lot of times that data you download it and it takes you time to get insights from that data so Finance managers can now use natural language with bedrock to access that data ask questions of that data and get the insights faster to share it with their team accounts payable manager and Auditors this this one is about documents and sap as we all know has a lot of documents involved in the processes within sap so with this process not only can you use we use SDK to um process the documents faster but you can use Bedrock to gain insights off those non-structured documents not only can you gain insights from those documents um that would have taken a lot of manual work before you can also take tasks or do tasks with the agents with that and that's a use case we're going to dive in a little bit deeper um because customers are taking that use case and applying it to other documents and other processes that they have in their system abop development um we have this uh abob development that we've released an abob assistant on our Solutions library and YouTube and with that we see customers using it to generate documentation if you were like me I had a 20-year-old sap system I had no documentation for a lot of my abob and I would change something and it would break so being able to generate documentation from my old leg code helped me would help me transform faster work on my S4 Honda rise drisk it and accelerate it so we have a lot of customers taking what we've built and actually applying it with different prompts to generate different kinds of documentation like test scripts as well as functional specs as well so it's just a great example of how customers are using this for their challenges and their benefits product development s similar to the Adidas uh that zot mentioned earlier you can explore product designs and Alternatives and then for shop floor instructions uh we have customers using Q business being uh their knowledge stores for Sops or training materials reducing their downtime by training their employees faster and they can take actions faster on their shop floor using Q so let's look at the accounts payable manager use case in a little bit more detail so we picked accounts payable because everyone has to pay bills right but the documents that accounts payable managers have to review is a cumbersome manual process and um like other processes in SA P if I can gain insights off of those non-structured documents I can make decisions faster and then I can use Agents from Bedrock to actually take um take complex taxs and make them faster for me so this is why we sett on the coun payable manager but we have customers applying it to other types of documents as well so as an example here um this is what how you can interact with the data with natural language with your documents and this is part of a demo which I'll show you but on the left hand side the users asking the model to look for Trends or patterns in the different documents around accounts payable So within seconds the model comes back and highlights number of key suppliers or use cases that were purchased to give you insights on that non-structured data on the right hand side the users asking for invoices that were not paid on time the model in this use case comes back with not just the list of invoices but potential reasons and insights on why it might be late so see this is where we're customers are applying this to other forms of documents like in um invoice reconciliation error documents Warehouse damage receipts as well so what I want to do now is walk you through a little bit of a demo um there it goes so by background the knowledge base has already been created for this for this demo and once the knowledge base is created we can start to ask questions of it so in the knowledge base it shows you the data source the embedding models and the vector database so now you also want to select the right model right this is where with bedrock you have the choice of model so you pick the model that's right for you and this one was we chose Cloud 3 Sonet so as an auditor you want to now ask questions of all this data that you have in your knowledge base so you might want to ask which invoices have been paid on time so what happens is that prompt that question becomes a vector goes into the database the knowledge the knowledge base does a a search for it and then retrieves the answer back to you so not only is it going to retrieve those answers back in seconds it's going to show you the source of the of where it got that information so it gives you a lot of information back to it and then you can also ask a question like you have a specific vendor that you want to find out about and you can ask a question do I have any uh invoices for this big truck repair shop so you you put that question in and again the model will take that go back create a vector search for similarity and come back with an answer as well as a data source for that so this is just one use case to show you what's possible how you can innovate your processes your data and gain insights on your sap data and your documents and again we have a lot of customers in many different Industries applying this use case and we actually have a session later today a workshop that you can walk through this Hands-On we have a YouTube video and a Solutions library that we've published on this so you can start to look at it as well great so in addition to getting insights from non structured data we have a lot of customers wanting to get insights from their structured data as well so over the past 25 years we've done a lot at at AWS and Amazon with AIML and we've learned that data is critical your data foundation in creating that modern data architecture is critical to get insights out of your business so we have many tools to help customers set up that modern data architecture that will set you up for Innovation and transformation when you think about generative AI your data is your differentiator and the reason is because we all have access to these great tools and these Foundation models but getting business value for your business is where your data will differentiate you in your process so we chose accounts payable manager you might have a different process that's more unique to you that you want to have a challenge with or you know you're going to have an opportunity with with with bedrock and Q you can bring your data use your data and differentiate it to innovate that's most important for you so a lot of our sap customers in addition to using Bedrock are also using Amazon Q so the one example I gave with a shop for employees being able to use and search your knowledge stores across all these different areas being able to act faster reduce um the down time on their shop floor is one area the other area is Amazon q and quick site with Amazon Q it's your orig digital AI assistant and with Amazon q and Amazon quick site you can take it where you work and you can um dive into your data get insights faster and use those datadriven decisions to um make better decisions and make them faster and with Amazon q and quick site what's great Testament of quick site is we have over a 100,000 customers already using Amazon quick site so with Amazon quicksite um it has a ml gener generative AI power behind it and what you can go into gener to q and quick site with your sap and non saap data and again ask questions in natural language anguage what's great about it is you can ask a question like how many sales orders did I have last month and Q and quick site will propose the visual for it it will propose what kind of graph works best with that data but let's say you want to Stack bar chart based upon my orders by region you can ask that specific question and it will return the chart that you asked for another way that you can use it is if you have a quarterly Business review or a monthly business review and you have all this data that you have access to you can ask Q to give me me a visual storyboard with many different graphs asking what changed in the last month of my business you can then share that out with your team and you think about the time and productivity this would save being able to ask questions and interrogate it um with natural language we see a lot of customers sap customers as well looking at Q and quick site to gain those insights faster one of the other things we've developed and I think we have some workshops on it or sessions is um some accelerators in the OTC and the procur to a areas for sap data so what we did is we U wanted to extend that to q and quick site to give our sap customers an idea of how they can use their sap and nonsp data for OTC and procure to pay and que and quick site and what they could do with it what's the possibilities of that data in Q and quick site what kind of dashboards they can produce so again we have a solution a guidance that we've published on this we have some uh QR codes that Ogle will share later where you can go in and learn this in more depth and then schedule some time with us and we can go through it as well so hopefully these different use cases what we're working on with our customers give you some ideas of how you can innovate and transform your business with generative AI from AWS and sap and how you can differentiate that and take it to your business so I'm going to turn it over to Olga she's going to walk us through continuous innovation some of the use cases we're collaborating on with our partners and different ways that you can get started and how we can help you on your Innovation Journey thank you Beth inter thank you very insightful lots of use cases I think you just gave us a glimpse of some of them and things where you can get started uh but if you're still asking why generative AI why all these use cases the art of the possible in my opinion this is part of a of a much wider mindset and is the mindset of the continuous innovation so in today's business landscape you cannot be successful you don't have the ability to adapt to the business very quickly and this is something that uh starts with a transformation Journey uh the it strategy is just a part of a transformation uh and if you're an sap customer you might know about the uh clean Core approach and that is what sap customers are using to be able to adapt and transform quickly to new needs new technologies new possibilities the clink core approach allows you to avoid uh over customizing at the core and uh avoids uh being uh slow in upgrading so you have a clean Core approach and then you can use sap btp Services AWS Services the SDK solution that Beth and sulat were mentioning partner Solutions or industry solutions to adapt your environment your sap environment uh to any need that might arise we have very strong partners that already supporting customers accelerate the adoption of generative AI Solutions some of these Partners we have here uh are already using our solutions from AWS uh to develop uh industry solutions for all of you we have IBM using uh Amazon Q to develop preventive maintenance trust verification and in verification Solutions HCL is uh creating a manag uh water system that can provide insights uh based on generative AI information and provides uh fast engagement with your stakeholders and Data Insights and DXE uh is also using predictive maintenance we saw at the beginning as one of the use cases uh to provide um very insightful information and intelligent asset to your customers we also have deid using Amazon betrock in many areas uh for example in spend management we talked before about accounts payable but they also uh create in accounts receivable applications and category management and PWC is using training and infen also to build business applications in a sap environment and Landscape looking at all these possibili ities and use cases AWS offers you a wide range of capabilities uh to start or contining your journey we have been talking about the AWS and sap collaboration we've been working together more for more than 15 years and uh sap is using Amazon bedrock in the generative AI uh Hub you have access to the model now also including Nova uh through the btp applications uh from AP they are using training and inferentia uh to develop their own models uh and we have a very close collaboration in the generative AI space security is number one priority for AWS you've heard that many many times this is job zero for us and you have access uh to endtoend governance capabilities security privacy and responsible AI for generative AI applications AWS offers you also support to build Solutions uh with different teams to generate your pocs or help you scale and go into production and we bring the history and the experience from Amazon using AI ml uh Innovation for more than 70 years uh and we have more than 2 million customers in the database analytics or ml Services we've heard about Bedrock many times Bedrock gives you full flexibility h of models you can choose the right model for your use case you can change over time you can uh compare and Benchmark models so that is uh that is one of the capabilities that AWS offers you in your journey and it's very easy to customize your use case and your generative AI application using those models with fine-tuning rock capabilities and using agents for your activities in summary the sap and a collaboration gives you a lot of opportunities in your generative AI transformation and you might want to say okay how do I get started if you haven't started in your journey we think there are three key elements when you start thinking about your generative AI applications first uh we hear a lot from our customers that they want to start as quick as possible because the technology is there and the Executives the customers uh the market is looking for the next thing that you are going to bring with generative AI here one of the key elements is to Think Through Your Best use case you know your organization you know your company you know which use case is going to fit your uh needs and that might be very different from one organization to another so have a think through your own processes your own uh stakeholders and use uh use that information to define the use case that is relevant for you we heard as well that your data is your differentiator so think about what data do you have available and how can you maximize their potential and again generative AI is only a part of a much larger larger transformation so make sure that this is not a Tech play or a technology use case make sure that all the stakeholders in your organization are involved from League Finance public relations whoever you might think uh that is very important to have them engag from the beginning we talk a lot about customizing and it's not only about customizing your data it's also understanding your use case customizing the models and AWS offers a lot of opportunities and capabilities to do that and make sure that your data is ready of course we saw that in the in the iceberg so the data is the foundation for any generative AI application that you are building we heard here there there were a couple of uh participants already scaling into production and of course when you start scaling into production you want to measure the value in your business so you want to understand what's a return of investment to the application that you are creating and not that you are just creating something for the sake of the technology so make sure when you go into production that you have the right metrics and the governance in place to be able to to measure and correct if your application or your uh Your solution is not is not delivering the value that you expect as well when you are uh scaling the cost is a big factor you want to have efficient uh applications efficient Solutions so make sure that you also have governance and monitoring in place to monitor your cost and make sure that uh your applications are are efficient and again manage your risk be compliant be responsible AI that when you go into production is is a key element at AWS we offer you a lot of options to help you in this journey from prototyping we have uh an innovation Center that can help you also going into production to define the use cases we have pro teams that can also support you in this journey so there are many teams in our organization able to to support our customers we have training options as well you can start assessing your skills where do I start what do I know what I'm still missing what are my gaps and make use of the different training and certification options that AWS offers and of course you have the partners we gave a couple of examples of partners that are building applications for all of you and you can take advantage of of a large population here for AWS Partners additional QR codes uh that can provide additional uh resources to get started like the AWS Solutions library for sap the workshop of transforming your business with generative AI YouTube videos uh the AWS SDK for sap AB up so I'll give you a couple of minutes to scan all of them see people still taking pictures lots of information here moving on uh and of course if you're starting your S4 Han transformation Journey you can also schedule a complimentary rise with sap on AWS Workshop or visit our kiosk at the Venetian in the AWS Village in the Expo where we have also stand for sap on AWS with colleagues and experts who can help you here there are a lot of sessions at rain vent for AWS on sap uh we have a strong collaboration here so here are just uh the ones if you're interested including uh the workshop that Beth mentioned before uh that is happening later today with that I thank you all for attending today don't forget to give us feedback uh and we are going to stay here for questions uh because it's a Sil session we will just be here down downstairs for you thank you all

2024-12-24 08:50

Other news