Session Information:

Automation at Bosch: Scaling to AI – a practitioner’s view

The question is not if you automate, the question is when, how and how fast. Bosch Service Solutions applies its automation in various fields. From back office to front office. From scaling automation to AI within a customer centric organization.

Description:

• Experience the journey of automation at Bosch Service Solutions with:
• Who we are and why does automation matter to us?
• Automation journey thus far
• Practitioner’s view: The Bosch Service Center a global and omnichannel approach applies automation

Session Transcript:

From Frankfurt in Germany, to the world.

Patrick Sprites Sir, is joining us, he's the global operations, the co-ordinator of Quality and Automation Expert at Bosch Service Centers, since 2014. Patrick has been an integral part of the Bosch Service Center, a global team that focuses on enhancing and Customer Experience at Bosch.

In his role as Global Operations co-ordinator, he conceptualize and deliver it on harmonized processes and competence enhancement ushering in not just higher efficiency, but also improved process effectiveness equipped with a deep understanding of operational realities and transformation potential's.

Patrick has worked to identify innovative automation solutions and concepts to optimize and enhance the overall customer journey and operational efficiency bit with robotics process automation, chatbots, cognitive automation, or conversational AI.

Patrick is going to share with us a very practitioner zeal of how this technologists are used and implemented to create value. Patrick it's a real honor and pleasure to have you with us today. Very excited to have 2 to 2 witness your presentation.

Yeah, Good, good, Good afternoon. Good evening if you would be in Germany. So, so, thanks for having me. I'm looking forward to the next, roughly half an hour. So thanks for the short introduction, really, like I said, looking forward.

So, what I'm going to do, and it's just just say, just introduce me real quick, I mean, it's Patrick with name, like, like you heard. I would really like to focus in the next couple of minutes on three different pillars. So, I divided the presentation into three parts.

So, first of all, we're going to start and we're gonna have a short and rough overview in terms of, what are we doing within within Bosch, will be coming from, to set the context right then? I'd like to give you some insights into the automation and activities that we're doing with. A clear use case full course, and I'm gonna share our experiences that we made over over the time, or over the last month, and years, and the push service center.

So, combine at the end different automation technologies.

And, of course, this, as this is a common thing, I'm really looking forward, also, to your questions and later on. So, and I hope that you can enjoy the presentation. Right.

Well, we'll be coming from at the end, Osh established itself as a leader of IOT economy and IOT company, and recently launched the three Es tragedy. That means, we are an Bosch focusing on sensors, software, and services. And in push service solutions, as it's written into name, we are focusing on the service part.

Um, that means, at the end, we are delivering customer experience solutions, focusing on different parts of the life, took to commit, basic fee of, improving quality of life, basically, and that could be from the, emergency calls. So, for example, if you stuck in a car accident, for example, and you need some help, and most likely one of our colleagues in the operation centers are gonna help you, essentially, an emergency over, or at least verifying that, that, that you are fine.

And that you don't need help, Or, for example, that, if you're moving ahead in and you're losing some like it, just for, for big air carriers, for example, you could, could be in contact with our guys.

On the other hand, you're also going to see things, if you're stuck in an elevator, for example, I need some help.

Um, you most likely also get in contact with our, with our teams. These are some examples that we're doing for external customers, but we're also delivering services for Bosch as such as, we are focusing holistically on services.

And later in this presentation, you're also going to see this because I'm going to focusing on some services we're doing internally where Bosch at the end results, the customers such.

So, this, in a nutshell, what we coming from. The question in this presentation is obviously, why automation?

And we're always trying to, to describe this of leveraging the power of B, because at the end, nowadays, we are living in an environment where we live really, in an on demand, next day, delivery, generation, where people, or customers, would also rate, accompany them according to bad customer service. And bad customer service, that could cope with higher delivery time, that could be really longer waiting time, not immediately, not self-service. So it's really important that you take the customer experience in the account.

And to do so, you might need to reduce process execution costs because due to high costs, some processes can't really be done or needs to be done manually.

And if you're talking about, for example, a 27 service or 27, yeah, omnichannel multinational approach, then this could, of course, generate high costs, and therefore, you need to find a balance in-between.

And the third element on that, is the business model innovation.

I mean, with that, the following So just take, for example, Google Maps as an example, Google works in a way that it transforms basically, the business model, behind Classic navigation, 20 years ago, Location notifications was really, really difficult, but, because people were using physical maps, for directions, for example.

But, nowadays, this has totally changed, because that simple example of navigation, is nowadays, enhanced by artificial intelligence. It's enhanced by machine learning, for example. And that shows that with automation.

You can, um, shape, and you can improve your business models, and enhance and buy innovation.

And if we want to look in to the crystal ball, and if you want to look into the code, where at the end technologies, could go ahead, and where we're going with the existing and given technologies that are available today. We're going to see that we have two main factors there.

First is that we're probably not going to have topics that we're doing today in our tasks sector and an execution execution list tomorrow. So that means, by 2025, we cannot have automated processes, which we even could not think about that today.

And the other one is that, up to 50% of the tasks, which we see in a service related sector, will be ultimatums.

And this is, of course, a strong argument, just say, OK, let's focus on automation and look at automation as a whole, and this is also something that we did within Bush Service Solutions.

Because we understand that, we need to check that on an more holistic point of view.

How we did that?

So, for us, it was kind of a key, I would say dilemma, not really, that negatively free a phrase, but we had the task to first of all, define what automation means for us. And automation, of course, can be everything. So we had to slice this to include automation in our service offering. To be, of course, competitive in the market, to be, to leverage efficiency gains there. So, what we did, we, we actually slice this into three different pillars and starting on the left side, basically, with the traditional traditional RPA stuffs where we're doing really repetitive routine tasks using RPA.

Um, Moving ahead in that, let's say automation chain, I'm applying automation where we want to involve, at the end, computer vision things, A basic language understanding, too, Arrive at the end sooner or later on the artificial intelligence part, too really.

Mimic or enhance human capabilities. Such as voice, voice. Automation texts, understanding things. Because we see that this is really, really necessary within our business organization.

Because, if you see that from an customer service point of view, you most like the understand that you're starting in the back office. And then you mature, more or less to the front office interactions. So, having always, of course, in mind, increase of speed.

Increase of quality, reducing, ever changing times, and generally speaking, having this in a competitive way to service the customer needs the best way.

And ideally to do this on an holistic level. So, we're always trying to build automation as a toolset.

And based on the customer issue or the process demand, we are putting out or taking our tool out of our toolbox.

Um, and to start, basically with that journey, I just want to give you, give you the example of RPA and afterwards and moving ahead into the real world, where we, we, we, we are today in the Push Service.

So, as we have understood the, the why, and the what of automation, the question now is, how is that concerning the organization, and what is, what is the deal to talk about that? And, to give you an example of how we tackle that within us all, at the end.

We started that journey back then, in 20 16, where we were evaluating, basically, T approaches to, Of course, accelerate efficiency and future positioning on, on a really explanation of exponential scale, Because, for us, we really wanted to create and to apply technologies that is available.

That is turned into reality, to get rid from the slide to thinking, of course.

But, then, in the same topic, we wanted to, it was clear from the very beginning that we want to get rid of the proof of concept stage, to have something that is sustainable and can scale up afterwards. And it's not like a technology that you're just putting aside in your shelf. And you're happy that you have it. We wanted to really execute on that.

So what we did, we found it basically center of excellence also within Bosch to move ahead there to use Bosch wide synergies.

And actually we draw basically a process pyramid.

Where we differentiate that the different, let's say processes. And we wanted to make sure to transform the transactional process there.

two, to move ahead and to have some really efficiency gains on them.

And we also, of course, thought on how to measure that, and we came up with a two main KPIs. The first KPIs is actually the bot hours.

If you want, you also can translate that into our, given back to the, to the organization to utilize this time in whatever manner. Could it be really efficiency focused or quality focused?

Things like that, and the other one, cops, more or less with the democratization of automation, that means we wanted to build a framework where we can basically train business users or users generally to automate their own processes, so to really democratize automation across your organization. And, thus, we did, and in 20 19, we have 10, more than 220,000 bot production hours to give you, give you a digital era and the feeling.

Of course, it was not. The stopping party was just the beginning, and the next time, the next couple of minutes, I would like to focus on how we put that into reality.

Combining different automation technologies within the Push Service Center, which is, I think, a good example to show what is possible and how to how to structure that, so let's take the look on the journey. How we, how we went there so far.

So, the push service center, generally speaking, the push service center, is a global service that covers 86 countries. We are doing that for 26 languages, on really an omnichannel scale.

That means, we have call e-mail, so like the traditional ones chat, but we also have social media stuff. While we managing, for example, more than 90 Bosch apps or so, more and more than 75 social media channels.

Including this in more than 1, 1 minute interactions. And in this global global approach, in that global Omni channel customer experience. What you have to understand that we are, we are behind the posh webpages, that means.

Each end customer, who'd like to get in contact with robots are short, Big, Holistic company, goes to a corporate webpage, clicks, they're on contact, and wants to place a request, or in the best case, of course, getting directly a resolution.

And this is of course, a challenging part, because like I said, 86 countries is is quite something we need to cope with. That the language variety and the channel verity.

As we are using seven sites globally. We also have that let's say inter cultural, complex setup.

And within this environment, we of course also thought how to increase and how to build an automation accordingly.

And this is what I want to share with you there. one last remark on that of course for us and then the added value to towards ... is that we can orchestrate the incoming request flow and the customer does not have to follow basically the complex structure.

So we got rid from that many, 30 something drop-downs. That is, let's say, a customer journey nightmare.

So, we really focused on we need to do the Task Center homework internally and not bothering the customer.

And, of course, we saw, and we thought, OK. Let's move ahead with, with automation and see how automation can, can support us, basically.

And this brought us basically to the topic, first of all, taking one step back and see, do we have the correct contact channel portfolio fit, to even tackle that automation, or where to start with that automation?

And do we have a consistency, and in regards of data, do we have that, say, the basics and the groundwork that we need to deliver on automation accordingly, and to satisfy, basically the needs. And how we did that.

We basically split the digital workforce, how we calling this, and we split that, as I mentioned earlier, into really the back office tasks and the front office tasks.

And as you also know, in the back office task, we thought, OK, let's apply RPA.

And moving then to the cognitive part in the in the front office part, this, this was really, really important for us, because you have to understand in that complex world where we are. Our agents also need to cope with different backend system. Of course, be using one uniformed ticketing system and the CRM's and things like that, but still there are a lot of, let's say, look up options that our agents and our teams needs to deal with.

And this is, of course, waste of time and therefore we said.

OK, let's reduce these manual lookups via applying RPA.

Um two, have their efficiency gains directly on the agent desktop, which is a good thing. Because at the end, the agent is feeling that he is supported by automation and it's taking a bit the fear away that automation is stealing a job, for example. And this is something that is also from a business ethical point of view of something that we need to bear in mind also if we are applying additional and new innovative technologies as well. And then after we did that, We also said, OK, if we are optimizing it from a back end point of view, how to see, and how to deal with a, with a self-service point of view of how to apply cognitive pots, for example, In our ecosystem to cope with our demand 24 7 in some countries in increasing volume over time. But of course, usually budgets are not increasing. So this is how we did that.

So like I said, we put the RPA stuff in the backend, um, to enhance really all incoming tickets or immediately with the relevant information, being it, taking customer, customer information. And enhancing this with backend information that the agent and does not need to go into the system.

And checking this accordingly because this of course, at the end, is coping with average handling time reductions. And we talk here about up to 5 to 6 minutes. Every training time savings of poison in some dedicated cases there. And then of course, we said OK, let's move on, move ahead with chatbots.

Because there we saw some efficiency gains in terms of self-service immediately.

But of course, the question also there is, how to get and how to start.

And the first thing we did within that, let's say number data crunching, we invested quite some time to analyze closed tickets.

Um, and we had to go there and we did like 30 K of tickets that we clustered and tackled in English e-mail volume. Because, I mentioned, also, do we have the correct portfolio channel fit. So I'll come to that in a second.

So, of course, what we did there, we did like a classical matching between the topic between GPS. So GPU is basically our divisions to see where to start, where to focus first, and where to do the correct prioritization.

To identify, basically, the most frequently used things, and to deliver and designed, and also the content history, to, let's say, set up, back then, the decision trees, quote, unquote, carefully and correctly. So that means, generally speaking, get the use case ready. Do the data deep dive before that you do your, your prioritization, correct.

Now, how, how did we do that? I mentioned that from a channel, and from a portfolio fit point of things?

We started with, with English volume.

And we have, or we saw the issue that we have an increased volume over time, also an English call, an e-mail volume, And we had no idea at the end how good automation could work, and if that fits our operations, our ecosystems.

So, we had to give it a try it.

So, what we did, we wanted to, to check, also, technical feasibility in our system ecosystem. Because this is something that is really, really important, don't, don't start with with a standalone solution, especially if you're talking into a global organization that should scale afterwards. Because potentially, depending on your company size, it will come back.

because people, or let's say IT strategies, are usually quite resistance against exceptions. So make sure that this works already. So our prerequisite was really to have a technical feasibility. That that really, really works.

And then, of course, as we didn't try this out from the very beginning, the big topic, boys gain experience in data redesign.

Because we we had to do this number crunching before because we had no idea how good the data is to be used. And, we also had no experience, how poor teaching, and how things think this works.

The next issue we came up with was, you can't really do a chatbot, obviously, if you don't have a web chat.

And, traditionally, we were quite strong in, in call, an e-mail, but Web Chat was not a preferred channel back then.

So, so we had to come up with the live chat as a channel first, which is kind of a logic thing, if you want to deploy a conditional chatbot, obviously. The reason why we also focused on a Web chat, and, not, for example, Envoys, come to that a bit later, but, we, we saw that we have an increasing call volume, e-mail volume also in English.

And we have, like I said, a global setup and we work with a chat opportunity. We saw that we can leverage basically the bilingual agents. So, if you take a Nordic agent for example.

Um, that is easy because people speak fluent English with a slight dialect, but you can't hear the dialect really in in a chat conversation, in a call obviously. Yes.

And that could lead to some customer, let's say Irritations or a bad feedback's there. So we said, OK, let's, let's use this on a global routing, stretch the approach, and applying it in chat, and utilize our H and B lingual agents all over the place.

And in addition, to, of course, set already a fallback option up if a potential chatbot in the future would fail. And this is something that is really essential, because we always need to bear in mind that there needs to be a fallback option in place.

If the bot fails, for example, and that technology and intelligent automation is failing, I think, you have heard also over the day sometimes. So, you need to better than mine and tackle it that strategically, already from the very beginning.

So, but, if we did not do that, jump directly into them, that conversational chatbot point of view, because we still were not 100% sure how this, how this works. So, we, in, not invented, but we apply something, we call the knowledge. The knowledge, but as such, was, for us, an intermediate step, with, let us say, reduced FAQ skull, and we applied an embedded that actually in the context webpage, where people usually start writing some e-mails. So we, at the end, put that in the contact form, and into search form, because then we could play with that, a with the algorithm a bit. And B, we can have an eye on the customer journey, because two aspect, she was not expecting that an FAQ is popping up next to the contact form when the customers writing an e-mail.

So, we just try it this first to get some feeding some ideas accordingly, and then to move ahead, I'm also continuously developing, then the feedback that we saw in the backend, or in the and then, moving ahead and doing the step into the conversational flow, um, yeah, having the learnings from before it end and for us the prerequisites to do to step into the DT chatbot world was OK, we want to have sufficient amount of FAQs, we wanted to have the correct data in place. And, then, we also wanted to have the consistency, India, T integration, And, this is something, that I mentioned before.

There is nothing worse than an automation, or a chatbot, that, at the end, fails, the chatbot office, the human hand over, for example, or the customer requests a human hand over actively.

And then, the customer needs to expand everything again.

So that human hand over comes with different. That's a prerequisite as well.

So first of all, of course, the prerequisite is that you have a fallback option that you have a human in behind. That comes with some challenges, because you need to ensure operation of that you have a person in the back, right? Because you can't offer on the weekend, For example, a human hand over if you, if you're not staffed operation.

So therefore, we took the route of 24, 7 first to always ensure the human hand over, of course.

And the second prerequisite of that is that you are doing a proper handover of data that the agent really sees, what is in what was discussed before, and can, let's say, directly impact the customer journey because you can imagine that the customer potentially is either annoyed or distracted or frustrated. So, you need to also train your agents on a way to say, OK, if you are now stepping into this conversation, that means that you probably find a customer who's not really happy, so, please be prepared, and ideally come up with a solution. So, this hand over, and the transcript and that forwarding is really, really essential.

And that copes with the anti integration towards that omni channel communication platform as we are using it, which is really an important part to not lose and not frustrate the customer accordingly.

And, therefore, also, to cope with the customer journey, we set a threshold relatively low to have an immediate hand over to, to the customer, to, to the agent approaches, that the level of frustration is not getting getting too high.

Um, So, to sum this up at the NP set, the knowledge part is kind of the intermediate step that we are using also for, let's say activity data. redesign and putting done FAQs in the back end, NCQA always the customer experience.

If we are, let us say placing ourselves in a really customer critical point of view. Because, as you potentially also know, the customer journey and the customer service organization representing, basically the brand company, and, therefore, I set that also at the very beginning of the presentation.

During these times, a bad customer experience and a bad customer journey can really hit you badly, so therefore, this is really important to motivate everything from from the customer point of view.

Having to set, at the end, this is the transition tool, a bit of a of a summary and the lessons learned point of view. Let me just quickly re-iterate what we did.

We started on a, let's say, global point of view. What do we see in terms of automation?

Where, where do we apply Automation? And how do we slice this with an S O? Remember, the three different pillars that we have there.

And then be motivated that on a global project. And then that takes the local skills environment, where we are applying different methods of automation.

According to the neat, if you remember the RPA in Bush Service Center, this is more motivated from a support agent assist point of view.

And, on the other hand, we are talking about cognitive automation, if it comes to self-service options, for example, where we, where we continuously explore how to make the customer journey easier, how, how we can make people get connected to the correct contact, if you want to, the correct information according.

So, as the biggest lessons learned, and I've put the five, generally speaking, one important thing is, you should start small, but start, but started.

It's, it's always kind of a thing.

Frequently heard advice, but it's so true at the end. If you fail, then please feel fast, Because said, at least save your money.

So, really, just define your, your, your area of, let's say, automation potential. And just go there.

Of course, have in mind all of the stuff that I mentioned before, the IT ecosystem and so on and so forth, but it's always easier to convince people with a quick mock up with an MVP, with a quick POC or, or even a pilot, um, because then you have an entrance barrier lower, because you can show something that you have something available.

The next topic is, and I mentioned that in the previous slide now, now frequently really focusing on the customer's issue under customer problem, and on the process design that follows at the end, the customer need.

So, if you, if you are doing this, then you are able to, to, to, to, to Gail afterwards, at the end, the technology such as it's nice. And this is what I meant. And in our RPA Center of excellence example, technology should not be something where you are proud of, that, you, that you know how that works. It needs to be applied in the customer process accordingly. And there, I'm using that tool, Box example, always there are several technologies, in terms of automation. available in the market. But unique to be able to use the correct automation, on, let's say, fitting customer issue, and phrase and shape at the end of the process around that, and apply it.

Yeah.

And then, of course, if you, if you've started small, then, you have the benefits, that you did to showcase into small. But then, you need to scale, really, Ought to take the synergies accordingly to move ahead. So the automation does not stop at the end with the project itself. You need to take, then, the potential to scale, according.

The fourth one is, it's it's nice if you have a small POC, but of course, upper management, and your your sponsor is at the end, money driven. So you need to focus on the payback time. You need to bear in mind, where is the ROI, and, of course, the faster, the better.

Um, but in this case, please, always bear in mind that you need to cope with a change management process, and it needs to have management support.

Depending on how your organization and your company is structured, it's necessary that you explain why you're doing things, and you need to do that from a stakeholder point of view, because, of course, central functions might look differently than an operations point of view.

So, take this into account as well.

And Then, of course, most likely, depending on how big you are, In terms of your organization, you're most likely, not the only one who is exploring these, these, these topics there Expedite your journey again, speak with, with your colleagues, utilize the networks there because to re-invent the wheel doesn't make sense at the end.

If you want to really execute fast and you probably need to, yeah, relatively quickly exchange with with other stakeholders in the company, to move ahead accordingly.

And yeah to two. To get feedback and to get started, or to get restarted, depending on where you are in that automation journey.

Having that set, and looking at the time, I would conclude this now, and of course, I'm happy to discuss additional additional questions there.

Please feel free to shoot them and just say I'm going to stop the presentation. So, thanks for, for watching. Thanks for for listening and looking forward to the questions section.

Fantastic.

Patrick, thank you so much for a very practitioner oriented reveal of what you have done, what the journey has been for you and the team at Bosch. That's that's really fantastic. So I encourage the audience to continue to ask questions as Patrick and I spend the next 10 minutes or so, you know, discussing some of your questions. one of the questions that came up that, I'm not sure if you, I don't think your address that directly. So I want to make sure to ask. How is your relationship on developing you know? Specifically, RPA? Without naming any one, what is your relationship being like with, with vendors? In supporting you, in that journey, you know, did you decide on a single vendor and you work with multiple vendors? What are some of the thought process and criteria that you use for selecting maybe partners?

Yeah. So this is actually a good question. So we we within us or we are focusing on one when the at the moment but we of course explore that intensively.

So what we did, we did our, That's how homeworks before. We invited a different several windows in terms of giving them a use case, exploring them, doing hackathons to really get a feeling on how people work and help people how we can combine and collaborate with partners there.

And then, but then it was relatively quickly, yeah, that we need to get these expertise in house. Of course, applying. Technology is from a wind up, a building that expertise.

Because we want to have that knowledge to dig it a tool, like I said, out of that toolbox accordingly, because only then our clients believe that we knowing what we're talking about.

So, at the end, what we're doing, we're doing an RPA. Our internal stuff with our center of excellence, with our internal, let's say, process consultants with our internal developers and accordingly.

And then it really depends on the additional technology we are talking.

We always tried to, let's say, do with them as something really fast, execute something really fast, doing our first, um, let us say experiences.

And then, if we reach a certain threshold, accordingly, we want to have this in-house at the end. Because, like, I said, we're looking on a customer problem. From a holistic point of view, we always want to serve the processes and the customers in the best way, and therefore, the combination of multiple automation tools are, most likely require.

Very good, another question, that the Surface has to do with the blend of, maybe, of technologists that, that you may be using for automation, so you specifically talked about RPA, and you mention a few others, but what does it look like in terms of the different types of technologies that you're using? You can reveal that one more time for us. We know you have RPA. How do other technologies play on your intelligent automation portfolio?

Yeah, so, like I said, we split that at the end and a clear back office back office and front office. So, in the back Office, Part B, we are talking about really RPA stuff with, with the stuff that comes with it. Right. I am also reading out PDFs and then, let's say, more shoot that we should kind of thing or more. And then, and then, you ish kind of things in the back end.

If it comes to the front end, we were doing several topics. I mean, we're doing within Bosh but also externally chatbot topics.

We, of course, within Bosh have also our big resource research organizations to continuously explore about NL you Artificial intelligence topic.

And if it comes to artificial intelligence, we within us or actively looking for, let's say, also, artificial intelligence that can enhance the customer during a customer journey thing. And, for example, that could be could be with a call, but you can motivate that and apply automation within really the call automation or an Agent ... point of things, or you see that in text, text interactions where we, where we are doing our research, isn't there.

We most like the combine external partners, but also enhancing by internal technologies.

Excellent, excellent, very good. And the next question is an interesting one.

Lots of compliments and praise for, know, failing fast to succeed sooner is starting small and scaling. And the, so maybe it's a two part question, The first part is that, how long has, has Bosch been this journey again? How long has it been?

So the RPA world started, I think five years ago, so 2016 is what we started we in parallel.

started to explore cognitive automation, in a sense of chatbot topics with being a big organization, of course, a lot of junctions.

Um, so also, in terms of, let's say, window management, setting up a center of excellence and things like that. But just kind of also essential to, to scale later on and to utilize the power of the organization. Because what, we always needs to bear in mind if it comes to Bosch, as such.

We are really, really diverse, and you want to see if you're looking at the posh portfolio within a squash it, driven by the customer by customer service as such. But Bosch has, of course, internal shared services, we see that we have IOT domains. We have customers, or to see costs must be a B2B customers. So, there is really a big organization that you need to capture at the end, where you need Center of Excellence and this exchange. Because otherwise, you most likely gonna lose synergies and efficiencies because.

If you don't know what, for example, the guys in China are doing, or in the US. Oriente headquarters in Germany, then that is, of course, a waste of a waste of money, and time, So there's, of course, takes some time to build within a big organization.

But to come back to your question, at the end. We have started, like, five years ago, but continuously doing this.

And always Seeing and scouting in terms of innovative technologies according Yeah. And the follow up question to that is that you know, you have been on this journey now for for quite a few years, and you'll learn a lot of things in this journey.

What you, what you, what do you wish you knew five years ago that, you know now, that would be maybe an advice for someone who's getting started, and then started in the right foot in the organization. Maybe less and learn, perhaps from your own journey, that if you go back, you have done it differently.

OK? There's, of course, cause cause kind of a tricky question, because I'm as I'm wearing different hats. So. if it comes from a project point of view, and now I'm taking again the practitioner a view on that.

Sometimes it's easier to start to start a journey and kind of a submarine and say, OK, we have something, and we deploy that. We have this, and that result. and this finding before you are, let us say, captured in, let's say, central processes that are slowing you down.

A colleague of mine always said it's easier to apologize rather than to seek for permission.

That, of course, does not mean that you just should do that on a stupid, random way. But that means if you have from a project point of view, a need, and you have, let's say, a technology available, in our case, in a specific division.

Let's do this, because then, let's say, a central function wouldn't need to offer you an alternative.

And as we know, big organization, that could take some time, and you you have then at the end, at that time advantage, and that experienced advantage.

two.

Yeah, really, really.

You also could then change a platform where you could change, for example, a tool for, for instance, But then you have already your experiences on. Of course, we then can benefit also from, from, from the synergies you're getting basically from central functions I would say. So so so do something to generate relatively quickly and MVP and bringing it down on the streets. And the second. And last point is, B can be a bit careful with that, to not create a POC graveyard at the end because this is done on the other side of course also a threat.

If you're not reaching the scalability face, that could be a bit of a challenge if it comes to documentation towards management.

So, the summary is, start something quickly, um, and then try to scale, and look for the synergy partners that you also see in the organization, but not ask for a solution, and wait forever, because the technology is really, really fast developing, and you might be behind the market.

That's excellent, Patrick. Thank you for that. We have like NaN. I'm going to ask a very quick bap bap practical question that's coming from Natasha Eckhard on this one and natasha Question is 2 2 2 quick questions. one is that how many developers are in the Center of Excellence, and then, how are you measuring success?

That really depends, because we have to differentiate, that we have different organizations covering automation from, from a developer point of view. I think in, in, in these, let's say, Robo Squads, how we told that, while we democratize RPA, for example, we are talking about, let's say, roughly at 150, something like that. But this is within us all. We also have additional, let us say, solution. Architects are developers in different departments.

So, there the overall amount is I can't give you a definite figure there. And, the second question just helped me with that, what was the question there? Again, is really good wrap up on how how are you able to measure success, you know, what are the FBI who may be using measurements for success?

Yeah. So, of course, this depends a bit in which you are, but we within us all, we talk about our given back to the business. So, this is really the, let's say efficiency gains that we see there to see also the development over time. Are you able to increase basically the deepwater hours and that you can then easily transfer at the end with the efficiency gains? Because either you don't meet the people. Or you can focus your people on quality, on customer experience or things like that. So I'm always giving back to the customers basically one of the KPIs.

That's fantastic. Patrick, Thank you so much for taking the time to share your expertise, your wisdom. Very lots of great practical insights for the audience. I can tell everybody is very pleased that in and grateful for you to take the time to share that with all of us.

Thank you very much.

Thanks for having me. Have a good evening. Thanks.

Thank you.

Thank you, Ladies and gentlemen. That was Patrick. Sprites are directly from Frankfurt. He actually is in mines near Frankfurt, correct? And to the world. so great to have to have him with us. And I hope you enjoy his practical insights now. This was the last session for today.

We're wrapping up, they choose now and tomorrow we kick off another terrific session with four world renowned speakers sharing their expertise with you directly. So very much looking forward to what's coming tomorrow. We started the same time as the sessions today. And that will cover the areas of Digital Transformation, Artificial Intelligence, Robotics, Process Automation, and a host of other Intelligent Automation approaches, from a strategic standpoint point, but also a practical standpoint where we can derive those insights for value creation. So, looking forward to seeing you all again tomorrow, and if you want to make comments and get any updates on what's going on, just look under. My name is rose appears on LinkedIn, and placed her comment in there. We're gonna put some summaries there later on today. And we can continue the conversation, that, way, until tomorrow. See you all tomorrow. I hope you have a great rest of your day, wherever you are in the world.