 All right. Thank you for coming. Thank you, Mario for being here. I Never thought I would be talking about yarn and bobbins at an open stack event. So this is the first who knew Tell me a little bit about for those who weren't was everybody in the keynote here Everybody was in a keynote. No, okay for Quite a few weren't so why don't you give us a quick rundown of what what early con does? Because most people don't know what it is. So early con is a textile machine manufacturer We basically deliver big plans to our customers where they produce in the end produce yarn. So yarn is the Product where every textile fabric that you use in your daily life like those chairs carpets and everything else Is made from and we do this on a bigger scale So half of the world's yarn production is running on the machines that customers bought at early con may may fiber Put that into context. How much is half of the world's yarn production? The world's yarn production is 70 million tons per year and 35 millions are produced on the machines bought from us and they're bought from you You're not making the yarn yourself. No, we only do the build the machinery the factory We do all the all the things to enable our customers to do those things. Yes Now if I think about that industry just to add some more context here Is that a very forward-thinking industry is that we're talking about, you know People who really want to be at the cutting edge In terms of engineering we were always at the cutting edge Because these yarn things have very high Requirements for quality and everything so the engineering in itself of the machines always was top-notch You were gonna say that there was nothing else you could say of course And now we're also starting to dig deeper into the IT topics Also delivering world-class IT solutions now All right, so what what was said like if you think back 20 years I know you haven't been with a company that long but but what was that tech stack like maybe even 10 years ago? Like single workstations in a bigger factory lots of workstations Over time then there were some came in some service systems like that Everything was not integrated. Let's say it like that. Yes. And what were you using at the time? Was it Windows based or you are Windows based of course? What what What has changed Requirements change, okay big time. I mean industry 4.0 digitization big topics Especially in China also big topic Chinese government is a very heavy investing in digitization topics and so our customers are also pushing To get new solutions and at the end of the day, it's all about driving Up the efficiency operational excellence in those factories So yeah, they only wouldn't benefit on a commercial basis from it But also on new product basis and from the environmental perspective These are the big drivers. You mentioned China Asia is one of your biggest markets, right? So China is 40% of our revenue Yearly revenue and then greater Asia India and for carpet systems. It's mainly Northern America US and just to add some context to that. What do those factories look like? I don't think most people here have been in a yarn factory unless you have which I see some I see They work with you. Yeah, of course So the bigger installations are like 24 football fields in size and Producing daily producing something like 900 tons of yarn Which we will say let's say a train of four kilometers in length Keep it like that. Not bad Why open where does open stack have to do with all of that? I wonder. Yes, why Open-stake. Yes. So now we have these require new requirements. We have to do Deeper integration into customer systems integrate our systems better with itself More requirements are high availability in all such topics And then you wonder how can you? Yeah, fulfill those requirements and then you begin to think what you can do and this is where we basically started What are those requirements? Yeah Requirements currently our customers rely on multiple systems and Of course everything everywhere where you have a different system for a different task You have interfaces everything has to work properly If you have a more integrated solution, then you have a lower friction overall and better process integration and from this standpoint our customers want us to Yeah, to get on board workloads that currently are for example are done by SAP systems or such things To drive down costs again And so we have to find a solution to drive availability and other things are To drive the operational excellence for example our machine base machine learning based workloads And no process integration overall, which is mainly the biggest topic overall You mentioned availability, which I assume is a big thing for your customers What does it mean when one of those machines goes down for one of them? first thing they produce waste and lose money Because if you think of a big installation, you can't stop this factory producing Even if one machine in this factory where thousands of machines are standing is going down then raw material is still processed and but it's waste at the end of the day and this is Yeah, the most important thing to not produce waste Now you mentioned your Windows shop a Microsoft shop They've got a solution there that you could have used with Azure Stack and Azure. Yes First of all, nothing wrong using Microsoft from my perspective. So we are open we use basically what we Brings us the most benefit there We of course evaluated Azure Stack Was a natural choice to look at it But from a cost perspective economic perspective, it wasn't feasible to use it So we skip that was there because of the hardware or software It's a two-fold problem I think that they have the first problem is that they have very very high standards in terms of specialized or Dedicated hardware that you have to use and The whole setup is not as flexible as you have with this open stack setups that you can realize and The second part is that you always have this running license cost usage cost So in the end you decided on open stack. How did you get to that point? Did you evaluate anything else as well? What else? Well, that's the question. Yes So far our basic idea was we need a little data center in the factory So and then we said, okay, maybe a little cloud and then Azure Stack was out of the game and The next thing that we then found out about was open-stake was totally unknown to us just through research we found out about it and then dig deeper into all those Possibilities that were there at the end of the day. We had a you know some questions that we had to clear What were those questions? Oh? It's a long time ago wasn't January Just remembering you found out about open stack with a year ago Yes. Oh, wow. Okay. We actually started all of this end of January so in last last year around this time we I was investigating and Just simple things like what can we do with a virtual machine on this thing? And then how was all this honor all those network topics are working and everything? So this was the basic. How can we size it the biggest question we had because we Okay, normally we have quite big factories, but there are some cases where we have only one machine So we need something that scales from very very little to very big Yeah, and from an operational perspective, of course is always better. You have the same system basis So it can use all all the same system Regardless of the size. Yes, so given that you didn't know much about open stack. How did you get started? In getting answered the questions or doing the research with the question. Oh research I assume Google helped you but YouTube Academy But but in terms of actually getting started and doing a maybe a prototype project or something like that The funny thing was I doing YouTube Academy studies. I found a talk from a Talk was called operating a open stack cloud with a very small team. I said this looks good to me Watch the talk and when the guys begin to talk I immediately recognized great. They are from Germany and then I just wrote an email to them and said okay I know it's not your core business. You are a public cloud provider, but maybe you can answer some questions. We have So who is that public cloud with an AWS? No scale up It's a partner. We announced that the keynote scale up and Christopher's here. See you off scale up and yeah They were friendly enough and said yeah, okay We are willing to answer your questions come to Berlin and then let's have a talk And I see basically used him as a consultant. Yeah, basically like this And then I said okay, when can we come and ask in that in December? And then we made an appointment for January and I said what's the cost of all of this and he said nothing just come I said no, we don't understand this business We need we need to pay for this workshop So I forced him to write an official offer Yeah, yeah, is he getting checks now he seems happy there Some yeah So so you didn't do it yourself. Did you try to Stand up an open-stack cloud at all. I Looked into it. I Know I I just kept up this dev stack thing and everything, but I thought okay It shouldn't be only for me it should be directly on a bigger scale We have to Investigate it on a bigger scale with a bigger test system. So don't start with it on your own It's not only open-stack. That's new linux also for us new So everything new and then we said, okay, let's really talk to the people who know what they're talking about Did you consider any other vendors because we've got you know, it's for a suzer and red hat and Canonical and everybody else as well problem is lots of those vendors are not so prominent in Germany I know Susie is yeah, but from a discoverability standpoint at this time it was not so easy to find a company where you say, okay, I trust them in delivering the right answers to us Without going flying out or something All right, so You're working with scale up. You've got got a prototype running How do you put that into production? Putting the prototype into production What was that was a journey like and I Mean Better to begin at the stop. I mean we then after having the workshop. We agreed on Building some blueprints. So putting our requirements into some architectural paper And from this blueprints, and we said, okay, let's build for each of those blueprints a reference system a test system for our internal guys and for us to discover this thing and From there we know exactly did that and Used this as a basis to jumpstart our first customer project. Yeah, so Was basically like plans Putting looking at the customer sizing taking the right blueprint setting the system up Having lots of troubles in the meantime because you learn a lot Things can go wrong and I always will go wrong. So we expected this but Was no fun things going wrong What went wrong everything Everything doesn't count Simple things I mean you order three servers from Dell they arrive and no rate controllers in there So there's one week of your timeline gone Where normally the open stack installation should be set up customs Next thing then everything you tested spinning up in the factory You tested everything again those rate controllers didn't work. They knew about it So another day gone. So you have to fix something Some random Cisco images where switches to the port channels. Nobody knows why but we are investigating so all those funny things where Thousands of years of IT knowledge sit around the table and all of them say never seen that before None of those sound like open stack problems did open stack didn't give you any any Problems mostly not open stack problems There's something that we have to investigate around safe whether we currently see some Rocks because at one time it was in a read-only state on this note But the guys are optimistic to fix it But this was a yeah, it was a thing caused by this which failing. So we will see how this Will pan out the other thing is on the other hand We're not open-stack. It's not software related issues. So you I'm still trying to wrap my head around the fact that you went from not knowing anything about open stack to putting it into production in How many months nine months ten months eight months eight months. That's that's That I wouldn't have heard that three years ago or four years ago when I went to my first open stack summits I think is that due to scale up or is that due to the the nature of open stack today or what would allow It's a mix of course not only open-stake not only scale up not own. It's a mix of all the guys involved It's it's an effort joint effort and if not all the guys play together when you can't do it but Fortunately we could what did your customers say when you said we've never tried this But you know, it's just a few miles of yarn when we put in some open stack So our century the the CEO of century I mean, this is the reference customer from the video that you saw and just They make yarn for you know Nike and and you know footwear and the pretty high-end customers big sports brands high quality product. He does You we just played with an open open cards. We said, okay, this is brand new, but this is our plan This is what you get out of it on many levels. I mean for us open-stake is only the infrastructure so to say and enables us to to do all those nice things like content continuous deployment and so on where we want to go through go to but Especially directly on the infrastructure basis You see how some benefits that you can realize like a high availability for legacy system and something like that He saw that we played with open cards and then he said, okay We are the risks and then we he made plans together with our process experts What happens is something goes wrong. How many money do I do I lose and then said we are in one vote now guys so So he expects that something can go wrong But he is a fully on board and yeah, even from his side who gives his best to support us So what does what does your current setup look like you've got this first rest reference customer? What what does their system look like today that's up and running? They have our it's our middle blueprint. It's a three-note system that we can scale if we need more compute It's a hyperconverged installation that we are doing so Opus deck is running Control plane is running on all three nodes and it's running in Docker containers on there safe as five storage. I mean this is often the case KVM as a hypervisor and Then on top of this we are doing all those things for us And the good thing is that we have a complete remote monitoring and things like this via our VPN connection So he's always connected with our central systems and this was very helpful during all those catastrophes happening And what what what does that system actually do? So you're pulling in sensor data you controlling the machine. What what does that open stack system do? so the software we deploy on the system actually is called the plant operation center and It's doing all those things around the production that you have to do and all the reporting statistics You have shift changes all of those things. It's integrating with the customers ERP system Is doing the labeling of the ready-made bobbins? So if every few minutes you have some bobbins coming out of the factory, so to say It's not on there. What's the product is that he produced on this bobbin and so I have to label it It's a simple sounds like a simple process, but if this thing fails then You have a gain waste because after a few minutes. Nobody knows what's on all of those bobbins That does sound like a simple problem, but that's Sounds simple, but right big benefit if you have absolutely What about all those those IOT use cases and and sensors and the machines? What are you doing anything with that? Are you pulling in all of that data and using it there? Directly from the currently in this installation directly from the machines into this PUC system not Not directly so to say but we get the data in from our so-called guide systems The guide systems are the workstations that sit at directly at the machine and control the machine what the machine is doing And where you can somewhat reconfigure the machine for example, and from there we get directly get data So there's an event every time a bobbin is for example ready-made then event and data is Pushed into these PUC system and from there on other actions are called Now when I asked you about did you consider other? Options besides Azure stack and open days always the public cloud to in their systems Is that something you consider? Does that something your customers? Considered the cloud was a very bad word for our customers because they are afraid to lose Yes, some of their business knowledge in leaking data into the public And this is the most important thing for them. They need a solution They can trust and the best solution they can trust is the solution that runs in their forwards. They were scared of the cloud Yeah, exactly Is that industry-wide or was it just century? I mean we represent half of the industry in Almost all of them were scary of those word cloud and public cloud. So it's I think is representative. Yeah, of course interesting so You've got a connection over a VPN now that gives you control as you need But this is there's no public cloud involved at all and no, so we don't need to talk about Multi-cloud or anything like that or hybrid. It's not an issue If our public cloud our central systems and the edge clouds, it's some kind of the same construct But no nothing like a be a be a AWS or Azure something like that. No All right, so let's talk a little bit about what's coming up for you then You've done this This probably you've gone through this process. It went fast yet lots of issues What would you've done different? Well, let's go back one step. What we've done different now that you've gone through this What would you've learned and what maybe can other people learn that on the same journey? Done differently. Nothing. Don't work with scale up or something like that. Of course not. No, no, you can work with them No done sure done differently not much some little things Because if you don't make these failures or get these problems you don't learn from it. So For us was good. I mean, it was really a stressful time And we don't plan to Go down with the speed pedal on the metal. So to say so you've Standardized an open stock. That's what it's gonna be going forward Going forward. We have some big tasks that we still have to do to be honest as I said in open state is a good infrastructure choice for us and we have have We have it now in place more or less now in place We have to scale this infrastructure up so that we can deliver more and more dozens of installations per year The other thing is that we really have to Bring our new Common service platform to life on this system. This is the next big step. So it's a complete new software development effort that we drive With lots of services with all our machine learning Algorithms on there. I mean we had a little example the keynote already and we have much more to come In terms of machine learning process integration new services and the next thing that we will bring to century There's exactly something like this. It's a new service that will help him to visualize things like The dashboard like the dashboard thing from the factory really bringing this all over the place and in his factory and Even bringing some software development methodologies like the edge. I stand up into a factory. So it's yeah So so the yarn industry is getting agile. I said, I hope so Interesting But so what I'm getting at is that that's that culture Change that comes with that must be very it must be hard for you must be hard for the the industry you're working with I'm yeah, it's hard for us. Of course, we have to learn all of those things But when we see it's something good that we've learned we directly pass it on to our customers on the other hand, of course centuries a very thankful partner for this because he and really understands things like meantime to recovery and all of those things Yeah From this point of view, it's great But we also have seen when we have shown what we are doing currently with a century and what we are planning in the future from other customers who were more Hesistant in the in the past that they are also excited to get their fingers on that solution so Just looks good. What's the advantage for your your customer? Say it's where the sales guy so For our customers, of course for our customers We will drive the operational excellence in a big way because I mean today you are very reliant on people individual knowledge and In the future you can put this into a structure process. This sounds to totally boring. I know But in the real world in the business world those things drive your efficiency and this is what we are doing and Okay There are those machine learning things that also will come and will automate some some some things and So they will also benefit benefit from those things. It's driving all those KPIs in the right direction This is what our customers can expect with the reliability that they know got it It's gonna open it up for questions. Maybe for a moment. I've got a few more for you But maybe you'll see if there's anybody in the audience who's got some questions since we've got about 15 minutes left if you Want to take a little bit deeper? Robert, thank you. This is great These actually have several really short questions with really short answers if you'll indulge me So I want to make sure that I understand this you went from not knowing what open stack is to being in production on a Hyper-converged architecture that you built in nine months. Were you here? Just Yes, that's unreal. Okay, and you're using Docker swarm to manage your container clusters No, but you're using you're using Docker as isn't so what are you managing your clusters with the Kubernetes or underneath? So we're open-stack is running Kubernetes okay, all right And you kind of you acted as your own integrator on this project, right? You didn't hire somebody to put the pieces together for you and do the engineering you did this yourself No, no, this is the business standing. We partnered with scale up. Okay. I just okay It sounded your description sounded like it was more of a casual They've just kind of helped you out. I went from casual. Okay, but this is more casual and final pieces How large is your your operations team inside that runs this infrastructure for your customers? Currently we are using our the same guys that run our data centers internally. So no net new hires No net new hires until now. Thank you Robert took away half of my follow-up questions there. Let's say Well, Shawn Michael is walking up tell it since you were a window got one question for you Just because I'm gonna cut you off on your own question I know you talked a little bit about challenges, which is great if you could just give us a little bit of color on How you evaluated security because I know you mentioned your customers were worried about the public cloud But did was there any evaluation from your company to look and see the Security and any security or risk related challenges and then as a logical follow-up to that We're what have been some of the GDPR related requirements that you've had to deal with it. Thank you Yeah GDPR related requirements is We have some key principles the first principle is we do as a lot as we can at the edge so in the directly in the customer factory If we have to transmit data to our state of the central data Centers then we do anonymized data as much as we can Second thing is having an open book for the customer. So we always know what we transmit How often we transmit and how much and In our central data platform, which is it basically her doops system we have Systems in place where we at any point in time can say where is customer or personal information stored in the system in which columns In which rows can filter them can anonymize if we like to Can delete of course. So this is the answer to the GDPR question and security itself We of course had our internal guys who are specialized in all of those things especially the head of our data center and network team ran our internal GDPR Efforts for our internal systems Over a year. I think it was so he's experienced in those things and of course experience in all the security stuff with his with his team members and So they have their hands in there and what we are currently doing is Before we now scale this thing up and deliver dozens of system each year. It's doing a full security audit and try to get the ISO cyber security certificate for our system Enough Other questions Nope Are there any special requirements because you're operating in China? Yeah local Chinese data data protection Regulations regulations are in place. Of course have to also conform to these and For this reason, we will not only have a central system running in Europe where our headquarters are here in Germany We also will have a copy cloud copy for especially for our Chinese customer within China We are happy to have a production facility in China also to have already some data center server capacity there And this we will expand to have a copy of our central cloud system ready for our Chinese customer You said cloud there. That's a dirty word Maybe we clean it up Since you were talking about Being a window shop and you probably have a lot of legacy applications as well that You were running that maybe still need to run in those factories today as well. How are you handling those? With this new open-stack deployment. So these are planned operation center is one of the biggest legacy systems we have Even if it's still under the active development We call now called a legacy system because over time in the face of the software will change On the other side, we have all those guide systems that directly control the machine itself. So It's also running currently on windows But we are also have developments efforts there in place for some time quite some time now And there are thinking in those teams about migrating to a linux based base And replacing some of the old things there and at the end of the day We also plan to onboard those systems on the open-stack infrastructure Is this mix of windows and linux giving you trouble and running windows VMs Is that a problem for you? Just running the VM. It's not not a big problem, but I I would hope there was much more support for those enterprise Guys who come from the Microsoft enterprise Now we have our internal license server each time you spin up a windows virtual machine You somehow somehow have to get the license in there Everything should be nice like you know it from your enterprise data centers and I guess there is some room for improvement And we will see how we maybe can contribute there Well, let's see Let me just follow up one second before your question Contributing contributing to open-stack you kind of hinted at that there is that you don't have a user Which team but is that something you're thinking about? Yes Thinking about I mean currently it's too early to say yeah, we will do something in the in this community to But at the end of the day I said to our CFO once in a presentation and look at it like this Normally you buy from for millions of euros licenses from SAP and Microsoft now you get this stuff for free some consulting okay, but at the end of the day license of course is almost nothing and What about hiring two guys anywhere in the world sitting there contributing for us in the community representing us It's much cheaper, and then he said yeah, you're right That's that's That's kind of how it should be an open source Um Actually, I have a question that relates to that. I was curious being new to open-stack relatively As you brought your team here this week what you were most interested in learning at this point in your process and Also, if there is anything that you've heard so far that particularly piques your interest and for I had no time until today To attend any session only our keynote um Interest peaked up front by the Starling X project of course We have an eye on it naturally and we will attend some sessions on it and we will see if this is something All the other guys I haven't spoken to them in depth so More water it's the other questions anybody Three two one no questions I Got more You're right I have more water what you'll get a cold if you Robert you thank you You mentioned the consulting fees I would they're probably high enough that I would cuff to if I had to look at those but Are you thinking about moving away from that? Are you thinking about moving the operations everything in-house? Don't he's not listening and doesn't hurt him. He's got money. Oh, no, just I was open cards Thinking about how you can operate with scale up thinking about what we can do in-house Even if looking at other maybe other possible partners and looking what they can do for us I would actually want to add something that so actually it was our recommendation to for early con to Get open to technology themselves because if you I mean if you put so much effort into Changing everything you do to a different way then you shouldn't rely I mean, I mean, it's good for us But you shouldn't rely on on the single vendor out there are helping you you have to have the knowledge yourself So I mean and that's what they're trying to do. I mean to go to be true Why are you standing there? You might be able to answer this question as well How hard is it to find open-stack talent right now for you and and also for for you maria? It is difficult. I mean, we're trying to Essentially, I mean my company is in business for 20 years and we taught everything ourselves. I mean we learned I Know from scratch. So I mean that's also what we did with open-stack like six years ago when we started with open-stack And finding new talent. Well, I mean, it's always good to To connect with the community and there's always someone, you know somewhere and I mean, that's the best way to find someone Not only with in terms of open-stack, but generally what I see is that Because we're so higher developers everything else You just have to get to talk to people if they if you can talk to them If we can show them what we are planning to do and what we are already doing Then mostly they are excited to get on board But the problem is finding those people that you that you really want to on board Yeah, and that's not only valid for open-stack kids for everything else For sure. So you're hiring? of course, so If you want to have a chat, let's talk come to our booth 828 in the marketplace hall And we're not only hiring for our headquarters in Rammstein. We're hiring all over the world So basically you can live where you want and work for us And we built up a new digital hub in Munich. So Munich is your gig Then you also can live there and work for us. We will try to make it easy for you to come to Ehrlich on Yes. All right. Awesome. I think that's our time. Thank you for coming Appreciate it Mario. You're welcome. Thanks. Thank you. Thanks everybody