App Development & Automation with SAP BTP in 2025

When I set up the series back in December, little did I know how appropriate the opening in a world where change is taking place at an

unprecedented pace would become more than ever. Dupe businesses

need agile, intelligent, and adaptable solutions.

This is where SAP business technology platform in short B2P comes in. SAP B2P

is continuously evolving with new features and functions being added regularly to ensure it stays at the forefront of

innovation. Today you will learn how app

innovation. Today you will learn how app development and automation in SAP B2P can help you work smarter and faster.

A very warm welcome and thank you for joining our sixth and final episode of SAP B2P in 2025 an exploration of the

latest updates offered by the global user groups organization.

I have the pleasure of having here with me Thomas Fulmering, VP, head of product management for SEP BP application development, automation and

integration. Thomas leads product

integration. Thomas leads product management for application development, automation and integration in SE BP.

He has more than 20 years of experience in different roles in product management for integration, workflow and process management as well as business

applications such as SAP ERP materials management and supply relationship management. Before taking over his

management. Before taking over his current role, he was chief product manager for SAP build process automation.

Today his focus is creating value for customers through hyper automation and building a unified user and developer

experience with solutions like SAP integration suite SAP's integrated low code no code offering SAP build or the

newest pro code offering SAP build code and with this I would like to hand over to you Thomas.

Hi Tanya thank you. Yeah, I was hoping you would say uh save you saved the best for last, but you didn't. So, um hi

everybody from sunny Germany. Uh let me share my screen with you. Uh I

am honored and pleased that um uh I'm here to talk about uh my favorite topic which is application development and

automation. And um um so the question is

automation. And um um so the question is what is application development and automation with SAP BP? Well, it is basically pretty simple. It's called SAP

build. And uh what is SAP build? Well,

build. And uh what is SAP build? Well,

it's you know as we call a turnkey unified environment for AIdriven appdev and process automation. And that has a little bit of

automation. And that has a little bit of a history. We started with build a

a history. We started with build a couple of years ago. um actually

focusing on the low code and no code side of appdev and automation and last year pretty much

around this time we launched SAP build code as the proc code environment uh in SAP build and then we integrated it with

um with the ABAB cloud as well and now u I'm really super excited to share with you that we took the next step of actually you know integrating all kinds

of you know scenarios into Juul to make build truly your turnkey AIdriven appdev and process automation

platform. So the benefit of build is of

platform. So the benefit of build is of course then that you know it helps customers and partners and also our own SAP

applications to work across multiple development models. So be it the ABAB

development models. So be it the ABAB cloud or the SAP cloud application programming model or cup or low code no code if you will as a you know

development model and we do that by offering templates recipes pre-built content packages and now you know more

than 570 or I lose actually the the count and these content packages come from our own line of business applications they come from partners ers

and you know we also develop content there because it includes also a variety of SDKs. It includes uh bot templates

of SDKs. It includes uh bot templates and everything that we learn along the way what are kind of like the most um

you know prominent and most demanded kind of scenarios that we package them into what we call as templates as recipes and deliver them. And now

actually also with uh the support of generative AI so that SAP build can be used uh to build workflows and automations low code or full stack

applications extensions integrations into you know the SAP or nonSAP world and of course on top digital workspaces and of course we support with

SAP build partners for what we call as multi-tenant SAS applications and add-ons. So if a customer wants to build

add-ons. So if a customer wants to build on top of SAP, well here's the place like use SAP build. Now how does that

look like? If you look at it, you know,

look like? If you look at it, you know, a unified developer experience start actually with what is the central entry point so to say what's the central

access that you have for all your projects that you build um for app composibility or automation. So there's

a single entry point and that's called the lobby. In that lobby you can build

the lobby. In that lobby you can build applications with you know be it with the low code side of the house with build apps or with build code full

stack. you can uh build process

stack. you can uh build process automations whether you start with a workflow or you start with RPA bots that are orchestrated then in a process plus

also including business rules and of course you you know can build your business sites as yeah I call them digital workspaces and then when I mentioned it

of course also you have access here to an embedded store where you see you know all of the content package over here I filtered in that screenshot I filtered the projects that we have recently uh

launched for processes and process automation. See these are alone 157 and

automation. See these are alone 157 and you see from US US tax exemption certificates management um assignment of materials to network

activities treasury invoices projects purchase requisitions all of that stuff you find here and it's at your fingertip that you

are in this embedded store in the lobby.

You click on you know activate or create from template and then within a couple of seconds it's yours and you have it in your project. But build offers more than

your project. But build offers more than that. Now you can also in the control

that. Now you can also in the control tower of build and that is fairly new in a couple of weeks ago. This is where you basically as a user agree to the consent

that you enable genai and then you can make use of all the fantastic AI capabilities that we have recently launched and I will come back later to that. Of course, you know,

this is also the place where you have what we call as connectors in uh in the build lobby. So which connectors do we

build lobby. So which connectors do we offer next to of course the automation SDKs that we have for a long time? If

you say well you want to connect to actually uh Microsoft uh Excel or you want to connect to you know a third party system well there are automation

SDKs that you can use these capabilities there but we offer also the notion of actions and you see here um the actions in the lobby that are have been created.

This provides actually very easy entry point into the world of APIs. So you

just click you know create and then you can search all the APIs from SAP that are offered in the business uh API uh business accelerator hub plus also the

events that are offered there. So

standard things like sales order created or sales order changed you will find there and then you can use these APIs but it's even better you can also

connect this to your you know um to your integration suite if you use integration suite and you say well uh I have created already there on the standard API some

integration flows you can use these integration flows also with a couple of clicks as actions so that you can connect actually also from build via integration flows for instance to third

party systems. third party connectors are there, right? Plus also uh of course that you if you use build itself and you have cup cloud application programming

model and you expose that cup model as a service you can also search for the cup model and use that for maybe more comprehensive business logic. Okay, so

that's in a nutshell what SAP build offers there. Now let's come to what's

offers there. Now let's come to what's new, what's exciting and I mean everything in build uh is exciting to me but recently we launched Juul for

developers. What is Juul for

developers. What is Juul for developers? Juul for developers is a

developers? Juul for developers is a unified offering for AI capabilities across all build solutions. So be it in build code or build apps or build post

automation or also the ABAB environment the ABAB cloud we have now comprehensive AI capabilities for a plethora of of appliances I wouldn't call them use

cases but appliances for application development specifically existing code optimization and automation that is super exciting I will talk about this a

little bit later but also stuff like okay you create a unit test you create artifacts in build process automation be

it a rule be it a JavaScript or be it um a full process or a form you can use AI to to do that so it's assisted augmented

development what is also very helpful there's of course the next topic which is summarization which is using AI explanability so you can use AI to

actually ask explain to me you know this process with its subro processes and then Juul as your co-pilot that we're using will explain to you what this process is about and which kind of steps

are in the process or you know you can use also summarization on more complex decision tables maybe that you have in such a process so Juul will explain to

you what you see there and then basically you can start generating also new artifacts all with the power of AI but what's more you know than just generating something and we have that of

course also on the full stack app development where you can create code.

You can also optimize code. You can

build full-fledged applications but explain or use explanability also on the code that you might want to migrate for

instance if you are still on another platform or you use another kind of you know legacy system.

I will talk about these topics in this session today with you. And of course, we're using large language models, foundation models out of our ecosystem.

Um, and of course providing options to our customers and partners which kind of you know foundation model to choose. Um,

look at the generative AI hub. Um there

you can see and then you know we work actually with the industry leaders there and you know have diverse kind of uh large language models that we use but we also have SAP centric large language

models for instance the ABAB large language model which is very specific to SAP and it provides of course specialized

outcomes. Good. So how can this be used?

outcomes. Good. So how can this be used?

And we asked many and I always say a couple of years ago when we started investing more and more in you know AI capabilities and you know we have a history of using uh success very

successfully AI business services like for instance for document understanding the document uh information extraction service or data attribute recommendation. On top of that, we were

recommendation. On top of that, we were building recommendations, AI recommendations in our own SAP task center, you know, so that you get as a

manager, if you want to need to approve the financial document, you get not only the explanation but also recommendation um if you should approve the uh the

invoice or not based on uh the past execution and also with a confidence level. uh so you know these kind of AI

level. uh so you know these kind of AI services we used and when we started investing in actually you know augmented AI augmented development I asked

developers um you know will you use a co-pilot and years ago you know some of these developers were still saying well we are developers we don't you know we

have human intelligence we don't need artificial intelligence to do our job and of course that that that picture has dramatically changed over the last 12 to

to to 24 months. And why is that? Yes,

because you know the next generation of developers that is already in the market now you know expect that there is a code

pilot that you know helps and and improves actually the developer um efficiency and saying you want to

generate you can generate code make recommendations maybe not only with a co-pilot but also in line and then you know help with real time documentation

get data insights. It's more like really a productivity gain using AI than taking the job away, right? And then also recommendations. So for instance, an

recommendations. So for instance, an integration developer will be super happy if you know she can use AI for recommendations on

pre-built content, on APIs, on optimization of scripts, all of these things. So that's I would say developer

things. So that's I would say developer productivity.

On the other hand, business process experts find it still hard even after you know and you know my history at SAP is 25 years now and I remember that we were discussing how to model the

process, how to model actually process improvements in tools in the past and it's still difficult to model. Well, AI

can of course help. It can create these process models for them. Now

administrators many times administrators next to business experts have actually the task to um script complex workflows across multiple UIs for instance also

they are super happy if they have kind like assisted and augmented development they are next to the data steward who and that is super important for using

data products and you know we have of course a solution for that with the business data cloud you know to understand data to not only understand data models but actually understand data

and get more insight into that data. So

you see these are kind of like the high level efficiency gains that you can get with AI. And of course here with you

with AI. And of course here with you know um SAP build we provide all of that we provide the flexibility of using low

code and proc code methods be it with build code when if you look at this here you know you start creating actually a full-fledged application and you start

with actually the data model and the test data and so on so forth but you can also quickly build full stack UIs mobile applications with pre-built templates

and does this across different environments and the programming models like you know cap uh ABAB and of course also the ABAB restful application

programming model now that leads to actually a reduction of up to 30% and you know we have done a lot of research

on that up to 30% of application development costs by you know expediting application development in create

creating data models, app logic, unit tests and sample data with Juul and I think we will see in the next two to three years that number of 30% even

growing more and more and more right because even more complex tasks full-fledged solutions will be able to be built you know um across the different programming models and not

only simple single artifacts and we took the next steps steps already here because it's not only about creating code but also generating for instance

unit tests unit tests for ABAB classes core data services something that a lot of customers do not do because it's really cumbersome and that leads to

problems and now you can use AI uh to do that and um that helps of course tremendously I will talk in a couple of minutes about how what does that mean to

existing code like legacy code how can we use AI I there as well. Okay. But

first, let me go to the next thing that we just launched a couple of weeks ago that now it's there what we talked about for a while. Yeah. that you can use Juul

also to generate you a full-fledged process automation and yes it's a little bit of scaffolding of course because you have to work with okay let's generate

the process and then you know edit the artifacts and go in line and say you know use jewel then to create a decision

table yeah that you say you know I know kind of like the business uh rule and the policy and You type that policy in

natural language and what it does it will create you actually you know the decision table that you can use then as a decision artifact in the full-fledged process automation process you'll see

it's inline in the model but as well on the complete model of course you can you know use AI and use jewel there also to create for instance a data model a data

type for the process you can use that data type then and create a form and use it in the process process and so on so forth. And next to what I just mentioned

forth. And next to what I just mentioned as well, of course, we have this employee productivity when it comes to decision recommendations in the task center. I explained that already. And

center. I explained that already. And

the next step is orchestrate more complex automations with the help of AI agents. Okay. So now let's come to also

agents. Okay. So now let's come to also what is new and super exciting. we

launched Juul Studio and we launched at the same time actually uh Juul for Abab and that's super uh amazing because

we have of course a huge ecosystem of ABAB developers and they said well it's great that you use you know Juul for in build code for cap but how about Abab

now you can use uh Juul also to accelerate your ABAB cloud development yeah in ADT in the ABAB development tools for predictive code completion

like I mentioned before unit tests and explain also CDS models yeah core data service models and ABAB together and then you know on the road map we also

look currently at okay full-fledged ABAB cloud application generation with documentation chats assisted abup extensibility and so on so forth that's on the road map you can learn more if

you follow here you know the QR our code. Which products are supported?

code. Which products are supported?

Well, we proudly say nearly everywhere.

That means BTP ABAB environment, but also public cloud edition and private cloud edition as of 2025. Now, that is not the only or the

2025. Now, that is not the only or the only use cases. Of course, we also in ABAB enable that you can develop your custom AI scenarios with the help of an

ABAB AI SDK. So there's something that is uh popular as ISLM which is the intelligent scenario life cycle management in ABAB and there we provide

actually support with the ABAB AI SDK that you can build your own scenarios and you know build your own kind of AI support and that means we leverage all

the capabilities and of course the large language models that are hosted on the Genai hub plus of course AB uses the ABAB LLM Now what's on the road map for ABBA is

one of the most exciting things that I've seen because I've you know talked to the to the colleagues six months ago and I said oh it would be so good if imagine for clean core you could go go

as a customer or a partner and just explore the legacy code and you know have AI help to transform that legacy

code um and that custom code you know from the old world into the new world actually from ECC to S forana bring your

code into the cloud into ABAB cloud and that is uh what is currently uh still in the making we will be releasing that

very very soon that means that you can use generative AI to transform your classic ABAB code to a clean core and

ABAB cloud compliant code and That means you do not just do that with a push button, but you have explan explanations in the ABAB test cockpit so that you can

really truly use that that capability with Juul to understand what did your legacy code do because that is the biggest problem. I know that also 23

biggest problem. I know that also 23 years ago I was you know in a project and somebody had created actually you know um a body implementation with

custom code in ABAB and later on was 10 years later you know the person was gone had retired and nobody understood actually the well you know of course as a developer you understand the code but

it was very cumbersome. Now you have a push button that tells you expl exactly what this code does and it gives you proposals how you should adapt this code

and use it and use modern APIs and then you you can actually transform uh that code as well. Yeah, that's very geeky, very techy, but I'm very excited about

that because it will help many people.

Now let's come to what our customers tell us about using dual capabilities for developers. And you see there are

for developers. And you see there are Hankle, Dloid, Uniper. I think Dloid is one of the um the examples that I love to use because um you know the colleague

there Chip Kleinhexter said in utilizing Genai build code offers our joint CL clients easier integration capabilities.

A streamlined developer experience for cloud application development boosting productivity and driving tangible value.

And you see that the other remarks are also going into that same reaction. We

have really clear proof points that using AI with Juul for developers has a big impact on the productivity of our developers at the customer

side. Now what and because we get this

side. Now what and because we get this questions many many times is what are the use what are the use cases right everybody talks about AI and we at SAP

of course we know business processes really well and we said well there are certain use case categories that we structured in kind of like you know

maybe even in the future in a maturity model where we say well start here uh if you want to have a quick win But by and large there are you know five bigger

categories of use cases. The first is navigation. The second is data access.

navigation. The second is data access.

The third is system transactions, workflow execution and decision assistance. Let's start with navigation.

assistance. Let's start with navigation.

So navigation simply means AI helps users across different applications to get to the specific information that they need. Yeah. if

whether that's on you know ba user roles uh to provide training whether it's on navigating to the relevant fury app or

other SACP systems based on the user context and so on so it's navigation support the second is data access that means enable business users to retrive

monitor and track data through the co-pilot jewel and here you see there is so much in there like only if you think about, you know, a retail system. Uh,

retrieve clo clothing allowance information or you look at employee salary information, safety stock levels, supplier verifications, even weather

data for logistics. All of that you know AI can basically provide system transactions is meaning you basically automate the structured

actions of the execution of transactions within various systems. But it also means creating records like managing cost center changes, creating service

tickets in a third party system, post sales orders in S4. All of this can be done automatically. And yes, you're

done automatically. And yes, you're right. If you think, well, I can do that

right. If you think, well, I can do that now with APIs. Well, but what if you know a co-pilot can do that for you and you have the co-pilot at your fingertips and you just provide the data and say

create a ticket, create an order, right?

And the co-pilot manages all the complexity of the different steps that are needed to do that and maybe has even you know access to and conversations with other

autonomous agents. So workflow execution

autonomous agents. So workflow execution is also pretty straightforward. For

instance, gather input for price inquiries, create a business partner in a you know structured process, pull tasks or push tasks to people and if not

needed then the agent can automatically do that also for you. And decision

assistance is also quite straightforward. That means you know

straightforward. That means you know that you use AI for decision making providing the necessary data

recommendations. Yeah. and analysis on

recommendations. Yeah. and analysis on that data and say I recommend that you approve. I you know provide

approve. I you know provide recommendations on materials to reorder based on stock levels. I provide a summary of the analysis of revenue or

store performance. All of that is

store performance. All of that is actually a decision support and it helps business to act more flexibly and

agile. Okay. Now let's come to what does

agile. Okay. Now let's come to what does that mean? And if we look at these use

that mean? And if we look at these use cases also, yeah, because now we talked a lot about you use Juul as a co-pilot for code generation or for assistant

augmented development. I think the next

augmented development. I think the next level is of course that we use Juul not only uh you know for these kind of tasks

for developers. But let's take three

for developers. But let's take three steps back and say what is Juul? Well,

Juul is your co-pilot that comes from SAP where SAP makes every touch point that we have in our applications simpler and make every

touch point, every task simpler providing agents. These are application

providing agents. These are application specific agents. They come from our LOB

specific agents. They come from our LOB applications and this is what you know we have of course as the highest value proposition at SAP because we know these

processes, we know the data, we can provide you with these agents and then the next question is what if now a

customer a partner wants to extend these agents with their own jewel skills saying okay that's a you know an extensibility scenario like we have

extensibility scenarios. Also, when we

extensibility scenarios. Also, when we deliver applications, when we deliver user interfaces, when we deliver processes and workflows, customers and partners want

to extend them to their own needs. What

with what happens with, you know, the agents that we de develop. So, you want to create your own skills to extend an agent coming from from June from SAP.

And the next step is then when can I as a customer or a partner start building my own AI agents. Now let's look at this. So Juul

agents. Now let's look at this. So Juul

Studio uh has been launched already last year.

We announced it and we are just a couple of days away from you know a big announcement and the GA of this. This is

planned for this year couple of weeks and you will see that of course you can use Jules Studio as a part of SAP build

to enhance Jules's out ofthe-box capabilities you know from SAP to support your scenarios. You can create deploy monitor manage your own skills

and you can do that with the drag and drop simplicity of the low code no code tools. We will add also some pro code

tools. We will add also some pro code capabilities there because that also can go hand in hand. Yeah. And of course that also means that you can then integrate any third party system like

here in this scenario you know creating a ticket in service now that you can you know build these skills very simple and of course when you do that

you will have in the dual studio all of the you know benefits of SAP build at your fingertips. So let's say you want

your fingertips. So let's say you want to create an action to integrate with any system. Well, you just need to

any system. Well, you just need to upload the API specification to connect to the SAP or the third party system or you basically look at the API business

hub where all the APIs from SAP are and you can create that action and use that action as you know a skill in your in your dual studio. The next step is of

course I talked about the content. Here

it comes again. Well, we have created you know all of this content and it includes also different kind of content packages that you can simply activate

and then say well you know make this a dual skill that you can trigger a supply chain process. You trigger an HR

chain process. You trigger an HR process. The supplier invoice actions

process. The supplier invoice actions can be used to create the supplier invoices. the SDKs that we have there.

invoices. the SDKs that we have there.

You can use the bot capabilities, the automation capabilities there and so on so forth, right? And you can make use of course via that approach also with the

existing artifacts that you have created on your own as well and you have them in your library.

Now of course you can start then also creating workflows automations with all you know the benefits of doing that in that low code no code fashion here and

say well you know you want to quickly integrate these capabilities that you built from scratch into jewel for instance to get supplier information

here and create a dual trigger actually you know that you can use that that capability And of course it comes with transparency

and governance because also these dual skills like all the artifacts that you create in SAP build you want to govern.

You want to basically have a full-fledged control about you know who creates the skills uh is there workflow attached to that in which kind of

environment are they um yeah allowed to be operated? Like for

instance, not everybody in your organization will be able to create a jewel skill to integrate with an HR with your HR system to read all you know the

uh the annual compensation round information for instance. Yeah. And

maybe supplier data are also different from finance data. So you can freely configure that and you have full-fledged governance. Okay. Now what are use cases

governance. Okay. Now what are use cases for you know these jewel skills and I have just two that I want to share with you. The first is a jewel skill to

you. The first is a jewel skill to manage discounts quite straightforward.

What is the challenge? Well you know you need to determine the discounts for customers and that's usually time intense. It requires manual checks

intense. It requires manual checks across payment history outstanding invoices and discount policy. So it's

quite a cumbersome task to do and you know if you provide that as a jewel skill it's pretty simple because it requires just data integration. Juul is

able to analyze the partner data from S4 includes all the years of the partnership the payment history and of course also what is super important the unpaid invoices and we use actually here

rulebased logic. Yeah, where you can say

rulebased logic. Yeah, where you can say hey no discount for parties without in you know within unpaid invoices right discount ranges automatic approvals

triggers and so on so forth that is a a super cool kind of use case for creating your own jewel skills the second is procurement efficiency so we all know

the challenge well there's a lack of unified supplier portal causes you know that RFI RFP RFX delays you miss deadlines there's inefficiencies in the

pro in the procurement and the solution is well you use jewel inbuilt work zone and that will enable intuitive RFX event navigation and responses and you can use

also jewel to connect to third party supplier portals and we use which skills there so creating RFX events deadlines document changes and also you know the

the training data and so on that you have in works on there and connect that so these are just two use cases And now

we come to the most excite exciting uh part which is building custom AI agents. Of course everybody in the

agents. Of course everybody in the industry talks about agentic AI and we believe at SAP we you know have a really

very good approach to this. So you can design your own AI agents and that means

that you can um create agents that can master autonomously if you want. Yeah.

As far as you want autonomously without a human in the loop. Comprehensive

multi-step plans. What does it mean? A

comprehensive multi-step plan. Yeah.

that are these are activities that run m across multiple uh systems and there are multiple tasks

and maybe you know you also need to uh enable the agent on the way uh to

execute this kind of multi-step plan to be more um cognizant in that sense of uh

reviewing own work reducing oversight and streamlining the process. And that

means also that you can enable or should enable the agent uh to you know be orchestrated across multiple contexts

and maybe even um you know be integrated with other agents. And of course these agents will also help your process automation to accomp to to achieve more

complex crossunctional workflows that you then at that point in time do not need to model any longer as 15 subflows.

Yeah, that's of course one of the ultimate promises. Now what are AI agents? Of

promises. Now what are AI agents? Of

course, you know, um I think the big difference to if you look back into the past when you have been building um RPA bots, you had like attended and

unattended bots and they were like running on your desktop. Still these

were you know uh these were agents in that sense. They were bots that were

that sense. They were bots that were kind of you know modeled. They were like um pre-thought. Now of course an agent

pre-thought. Now of course an agent should be able to act autonomously and of course iteratively as well to achieve

you know a goal across multiple tasks and the key capabilities are clear. So

planning yeah the agent can plan multiple step steps reflection selfcorrection and reasoning different tools and APIs can

be used yeah to build that agent and of course what I just mentioned agents can call other specialized role specific agents and interact with them or with

humans right to achieve a task and there you know is the potential I believe it's huge that you know a lot of the structured and unstructured work,

structured and unstructured data can be brought together actually and be handled in the future by agents. Now, what are the use cases

agents. Now, what are the use cases here? Of course, you know, it's all

here? Of course, you know, it's all about automation again. So, you automate your business tasks and you augment and accelerate your business decision

making. And here the examples like for

making. And here the examples like for instance automate the creation of dispute cases from incoming collection mails right so you have unstructured data and then classify customer

complaints out of these mails and propose actions. So that's structure

propose actions. So that's structure classify incoming unstructured data.

Second use case is that you can automatically forward you know customer complaint tickets about invoices and then you know you augment

decision making by combining this information again so you bridge system and domain gaps there for cross domain automation that is also something that

is not without agentic AI that easy today right and now augmented business decision- making of course resolution of procurement clarification requests.

Yeah. Support human decisions by reiterating and reflecting on different data points and proposing interpretations. So that is like really

interpretations. So that is like really unpacking complex tasks to that level where you can say well there's a multi-step reasoning reflection really

understanding the process. There are

multiple tasks that needs to be executed. You get recommendations. You

executed. You get recommendations. You

can be in the loop as a human and basically trigger the next steps and so on. But that's basically that you that

on. But that's basically that you that an agent will be your next colleague so to say that helps you to achieve you know your business goals

faster.

Okay. Concluding I would like to show you a small video. That video is based on a really um cool use case that we

worked with the customer Freudenburg and our partner Svanta. And uh yeah, take a look at it

Svanta. And uh yeah, take a look at it and maybe this is also inspiring for all of you. Together with Freudenberg and

of you. Together with Freudenberg and Svanta, SAP applied this approach to a specific business scenario. Let's meet

Anetta, an IT professional that staffs projects with third party resources using PDF contracts. This tedious

process involves extracting daily rates from the PDF contracts, calculating costs and spreadsheets, and manually checking book consulting days in the S4 system. Using the business agent

system. Using the business agent foundation and SAP build, Daniela creates agents to automate these tasks. Inside the process builder, she

tasks. Inside the process builder, she adds an AI agent to her process. In the

agents configuration, she provides instructions and equips the agent with a document tool to comprehend PDF contracts. with a calculator and coding

contracts. with a calculator and coding tool for consultancy cost calculations with a web search tool to find further supplier information as well as an SAP automation tool to trigger SAP

automations that look up S4 system information. Let's have a look how the

information. Let's have a look how the agent looks in action via SAP build.

Danella can utilize a form to define her specific requirements for a consult. In

this instance, she's looking for which of their suppliers offers S4 HANA master data consulting services at the most competitive price within their framework

contract. Within a minute, Daniela

contract. Within a minute, Daniela receives an approval form with the chosen partner. She then reviews the

chosen partner. She then reviews the agents decision-making process. The

agent created a tailored plan for the request. It identified relevant S400

request. It identified relevant S400 consultancy contracts, ignoring irrelevant providers. The agent

irrelevant providers. The agent demonstrated understanding of each consulty's payment terms. And using its coding tool, the agent created formulas to calculate costs for each consultancy

based on a requested duration and country. Impressed by the result,

country. Impressed by the result, Daniela approves the proposal. Following Daniela's approval,

proposal. Following Daniela's approval, John, a procurement manager, gets notified. He verifies the supplier and

notified. He verifies the supplier and approves the purchase order creation.

Now, a robotic process automation agent takes over on behalf of John. It

automatically executes the creation of the purchase order by entering data and clicking options on the S4 HANA user interface. The AI business agent at SAP

interface. The AI business agent at SAP build process automation revolutionized Della's consultant selection process, transforming it from a time-consuming manual task into an efficient automated

workflow. This optimization allows

workflow. This optimization allows Danila to redirect her focus towards higher value tasks, ultimately leading to improved resource allocation and increased job

[Music] satisfaction. Good. So final slide um

satisfaction. Good. So final slide um from me today. Let's get started. So

here are you know QR codes for um how you can explore SAP build um try it and then join actually our

community and of course also explore our learning journeys that we have plus also more information in the SAP discovery center about SAP build.

Thank you. Back to you, Tanya.

you. Back to you, Tanya.

Great. Thank you so much, Thomas, for this excellent presentation. Um, perhaps

you want to unshare then. Uh, perfect. Thank you so

unshare then. Uh, perfect. Thank you so much. So, we had a quite a conversation

much. So, we had a quite a conversation going on in the chat, which is awesome.

And uh questions were answered by other attendees which is really very great.

Yeah, let me ask them anyway. So um let me start with the first question. So um can

we use Juul in sub build with rap or is it supporting cap only?

Well, in SAP build code, you can use Juul with cup, right? In the ABAB cloud as a

right? In the ABAB cloud as a programming model. In BTP, you can use

programming model. In BTP, you can use Juul with the use cases that I have explained. If you look at uh the slides

explained. If you look at uh the slides again, there's also a QR code where you can where you can see. Of course this is also the scope there is also increasing

um over you know the next quarters and we have a road map there where you can also use Juul then uh with the abup cloud programming model also

with look at the slide there is a QR code for the road map great thank you and um by the way the link to the slides will be available in

the description of this video that's a very common question Right. So, um the next question is, is

Right. So, um the next question is, is there a jewel trial? Is there a jewel trial? There is

trial? Is there a jewel trial? There is

a jewel or there is a a trial for SAP build that includes also the jewel capabilities. I think there's also a

capabilities. I think there's also a separate kind of jewel trial but jewel of course makes a lot of sense when it comes actually with the lobs or with the

kind of technology. Um you know we can provide you also here with a link to the respective kind of uh trial notifications or trial URLs.

Right. Thank you. Now do um I add to that of course we just launched um we just launched uh the genai

capabilities in Juul for process automation. I just have shown you that

automation. I just have shown you that there is in the control center this kind of consent and when you start using that yeah you get also uh in the first month

a free entitlement to try this out. Um

so that uh uh you should be aware of and every time you basically use Juul you see okay you get like a thousand entitlements and you have used already

200 or something like this. Yeah please

check this out as well.

Thank you so much. Now are there tutorials or examples based on rap?

Are there tuto tutorials and examples based on I'm absolutely sure there are.

and we will provide you with the links.

But uh that that question is very generic. Yeah, of course. RAP is a very

generic. Yeah, of course. RAP is a very common kind of part of the ABAB cloud programming model and they're for sure tutorials.

Yeah.

Okay. Thank you. Sorry, can more specific here.

Um and um is there a this connected to what you just mentioned actually is there a limit of dual usage per user per day?

Not per day. You can check this out also in our discovery center where we describe this right. I mean there is of course a commercial model on you know our AI capabilities.

Um but I mean there is there are entitlements that you can use for free uh in the first couple of months and then you know the uh the basically the

the pricing model with AI units and so on so forth will kick in. Um but that information is also available uh if I'm not mistaken on the SAP discovery

center.

Perfect. Thank

you. And now

um how can we measure and clearly see the actual value that SAP B2B brings to our business processes? And just to let

you know um someone kindly mentioned that um actually Jerome mentioned that it is possible via SAP cloud ALM.

perhaps you want to add.

But uh how do we measure actually the value the total the the the return on investment so to say you is

the question right and I mean of course we have studies and we recently have released also a study with a partner sorry not a

partner an analyst company uh called gigah ohm where we you know explored exactly that the question how much more efficient is it using BTP and

specifically SAP build um comparing to you know using hypers scale or using other technology and there were results like you know

30% increase in productivity and of course all of that increase in productivity comes with the benefit of a

return on investment but larger scale there's also you an ROI calculation where you can say well if you use this for your procurement processes what is

the return on investment if you're using BTP it's a mixture of course of developer productivity but also scenario return on investment yeah

perfect thank you um then there was a comment um you cannot develop rap projects in SAP build code or um business application studio or at least

not yet. Do you want to comment on that?

not yet. Do you want to comment on that?

No, that's not that's you still have you know the ABAB cloud environment uh with with ADT the ABAB development

tools and rap and we will continue to have that and they are also in ABBA of course specifically to the for the usage

of as in S4 um productivity tools there.

uh so key user tools actually there um but what we are working on is because the on the opposite on the other side so to say in in non-ab you have cloud

application programming that we are working on tool bridges that you can for instance you know from a rap business object basically continue with cloud

application programming and that this is a smooth kind of pool flow and content fluency from the rap business object to the cloud application programming uh and

plus things like common uh uh tools like core data service modeling across rap and cap. Yeah. Um so

you will not find rap now inbuilt code but more uh a unified user experience for developers across ABAB and non-abb.

Yeah, that's the direction.

Perfect. Thank you so much. Now um you already mentioned clean core um in your presentation but maybe the question here

is how are we supporting the SEP clean core strategy through BP? Which parts

have we addressed on stack and which ones side by side?

Yeah, that is a very good question.

There is a fantastic slide on this but I try to explain it just you know visually. Well, you have on the one hand

visually. Well, you have on the one hand you have basically you know the world of the application or the applications, right? You have like uh S4 private

right? You have like uh S4 private cloud, S4 public cloud and you have of course all the other cloud LOB solutions from SAP and all of them are a little

bit of course different but specifically on clean core the situation is that you have um for instance SAP ECC on premise

and now you want to basically move actually into the cloud and um that's one aspect of it. So the

transformational aspect of clean core how do you you know come to a clean core state public cloud by nature is clean because you cannot write custom code

there you just have the productivity tools and the abup development tools and you cannot basically write your custom code in there um and what I say relates

to public cloud uh sorry to private cloud and to on premise moving away from custom code and become becoming clean core is the story that I told you also

here in the presentation about first and foremost how do you utilize also the capabilities that we have in the ABAB

cloud um programming model not only in the cloud but actually in ABAB that you can for instance uh explain the legacy

code uh get recommendations how you actually change that code and make it clean core compliant. So that's

basically really on stack. And then

there's of course the world of extensibility and extensibility in BTP is basically a seamless tool flow and content fluency for extending your

applications then in a clean core compliant way with build be it like for app development with build code or be it with you know process automation for

workflows and automations and so on so forth. And that is what SAP build

forth. And that is what SAP build essentially is. You have access to all

essentially is. You have access to all of your um extensibility scenarios that you are building whether they are full stack

with Java, JavaScript or ABA cloud or they are with the low code and procode tool. So the extensibility solution to

tool. So the extensibility solution to be clean core compliant if you will though I don't like that that term so much but you know um to act in a clean

core way so to say uh is basically then SAP build that's the standard solution including the abup cloud. I hope that answer was not

cloud. I hope that answer was not confusing and it helps.

If any further questions come up um please post them in the comments area underneath this video and they will be answered. So let's go to the next

answered. So let's go to the next question. Is

question. Is yes. Oh no. Okay. Um is jewel enabled on

yes. Oh no. Okay. Um is jewel enabled on datasphere and sack or can we enable it?

What what what was the Sorry, I didn't get the is Juul enabled on the business data cloud and um no um data sphere and

um SAP uh analytics cloud yes or we receive this we receive of course this question a lot of times and I even have

a list yeah um because people people ask this I mean as of September already last year June is available Success factors,

SAP start, S forana cloud public edition, S for HANA cloud private edition, customer data platform, BTP cockpit, build code, build post automation, product life cycle

management, uh asset management, digital manufacturing, integrated business planning and business data cloud is just

has just being launched and yes, it's also there, right? Um and the list is growing every week, every month of course. Yeah.

course. Yeah.

Thank you. So the next um question is many

you. So the next um question is many times when we want to get these services enabled related to AI and Juul um this is challenged by our security

counterpart and it's difficult to proceed with the next steps. What

document steps has the P taken to make this secure and where can we find this info?

Uh that's a that's a that's a tough question here to answer just with a couple of sentences. I would suggest we provide actually an answer with a link

in the chat here because I don't think we have enough time to do that here now in detail.

Okay. Um, can you provide me the link and I can post it in the chat?

Yes, of course, Tanya. Perfect. Thank

you so much. If you send me the question again, we'll look it up. Yeah. Oh, okay.

So, later. Okay. Then I will post it in the comments area after the session is over. Yeah. Yeah. So, that's a little

over. Yeah. Yeah. So, that's a little bit beyond also my I mean I'm 25 years in the in the company. That doesn't mean that you can ask me anything.

Okay. I have to at times look things up.

I'm sorry for that. It would just absolutely no problem at all. Um we're

there to answer them afterwards even if not all questions are answered in the session.

So currently um there is no other question but um given that the um user groups have meetings actually they're

just meeting in world just now and we often get questions I would like to ask take the two minutes we have left as long as no other questions are asked in

the chat. So, how do you foresee Juul

the chat. So, how do you foresee Juul and Agentic AI influencing the evolution of low code and no code platforms?

How do you foresee Juul and Agentic AI influencing the evolution of low code and local platform? Okay, good. So you

know there is a discussion in the market and I had also discussions with a couple of analysts in there because there is the opinion that you know the

advancements that we are making in agentic AI. Yeah. uh that this might

agentic AI. Yeah. uh that this might have a potential to basically make platforms or

platform capabilities like for instance um low code no code or even RPA you know

um obsolete and I don't think that is really true. I think that specifically

really true. I think that specifically for low code no code the style of working will absolutely stay. It is just

it will probably change that you know you will not need to start modeling um a process or a data model from

scratch because AI is basically there to do it for you but you still want to have the visual representation of what AI does and then being able to then change

it again is also something that you want to do maybe even you know without AI or with AI. So low code no code will not go

with AI. So low code no code will not go away and specifically building agents will be much more dominated by

the by you know low code or even to a certain degree absolutely no code. Yeah.

Than writing code for an agent that needs to work seamlessly together. Yeah.

But there is even probably a market evolving in the next couple of well I shouldn't say years maybe months in with this rapid evolution here in saying low

code agent building or something like this. Yeah a market for tools for low

this. Yeah a market for tools for low low code agent building. This is how we see it and this is also why Juul studio is actually really at the end of the day of course for everybody but it needs to

work also for business users who want to build an agent in a low code environment right yeah great thank you so we are actually on

top of the hour there is one question I would like to quickly ask SAP uh Juul studio will allow the creation of AI agents or only jewels skills

There are two scenarios for dual studio.

One is the skills for extensibility and the second is build your own agents and both of them is dual studio and the other questions will be

answered in the comments area underneath this uh video in the next few days.

So let me thank you Thomas very much for your excellent presentation and answering all the questions from the chat and let me

uh express my big and warm thank you to all uh attendees for joining. I hope you got a lot of value from this session and

I hope to see you again next week. Thank

you. Have a bye-bye bye-bye. Have a great day and a great

bye-bye. Have a great day and a great evening depending on where you are.