 All right, I think we're going to get started so everybody's just getting back from their lunch break Yeah, that's part of the issue. So what are we gonna do? Have you do jumping jacks or something? So you don't fall asleep right now All right, we're actually really excited to be here today we're from Allianz, Germany This is Ludwig Paulsen. He's a project manager product owner for our cloud ops team And I'm the head of department for that same team and What we want to do is talk to you a little bit about what our experience has been Just making you know embarking on this journey in into these new Technologies and I wanted to also say that you know, we've heard a lot of deep technical conversations Throughout the last couple of days and one of the things that we have found though is that it's not just About the technology we'll talk a little bit about the technology But it was important to us to also talk about all those other puzzle pieces That it takes to make the digital transformation for an organization successful So who runs the cloud? Who runs the cloud at Allianz, Germany is cloud ops? That's the team and yeah, this is our logo and The team is who made the journey possible the team started as a very small team And we'll tell you a little bit more about what that experience was like But first I'd like to tell you just a little bit about our company the Allianz group is a is a large insurance company We have about 88 million customers worldwide 140,000 employees and We are in about 70 countries around the world and One of the challenges that we have I mean, it's a good thing and it's a bad thing We're 127 years old this year So the bad thing is that that what that means as most of you can probably imagine is that we have Lots of IT debt many many legacy systems But the good news is that we've always been at the forefront of innovation We made a choice back in the day to be one of the insurance companies with the Titanic project Didn't turn out so great But we're hoping we're more successful with these new endeavors, you know with Providing drone insurance and other exciting things so to talk to you about The the journey that we started a couple of years ago and and how that came about we had some of our executives visit a bunch of different companies at Silicon Valley and One of those companies was pivotal and they came back and what they said was that's what we have to do We want that you guys have to make it happen So so that's again a good thing we had senior management buy-in they were all excited about the possibilities that would come with that and So we started you know and it was in 2016 we had By March our first PWS Install we went with PWS because it was the fastest thing we could do But realized by by the time August came around that we really needed to do a little bit more and and Actually set up our first cluster on prem on prem because we had always done it on prem, right? And then by the end of the year, however had to really look at scalability of of Our environment and moved into a multi cluster cloud environment In parallel so what while all this technology stuff was going on we also Set up training centers to get our developers Developing in more agile ways we had sent a couple of teams to pivotal labs in London where they learned how to pair program and and also to use the technology and Then by the time they came back. We had to have our platform as a service our cloud Not cloud yet, but our platform as a service ready so that they could actually begin to develop again, we realized that that really didn't scale enough if we were going to be serious about digital transformation and So by the end of that first year we sat down and talked about different options and Just this incredible thing happened as an insurance company We always said we were never going to use the cloud, right? And by the end of that year, we made a decision. We looked at all the options and and We're able to decide to move into into the cloud. So we don't have data in the cloud We still today do not have any data that we store in the cloud, but we were able to move our front end systems in into the Amazon cloud But how did we do that? We learned from from that first year project where we had set everything up on prem and Had run into just a few issues especially toward the end of the project where You know security folks came up and told us about all the things we couldn't do So that delayed us quite a bit and so that when we decided to move into the cloud What we decided to do was to actually have cross-functional team a truly cross-functional team with someone from from our architecture team, but also from the security side of the house actively participating in the project and And creating the solution together with us rather than you know having having them be the The showstopper with the stop sign at the end and that worked really well for us. I have to say so what we did was we had those The new environment set up within five months at the end of that second year We were in the cloud with 600 hosts 11 container platforms 3.5 terabyte of memory and about 500 users using that platform and Out of that we had five apps in production and about a hundred in in our development space and Now a few months later. Do you know how many? How many we have about a hundred applications now, so Cloud Foundry alone is 600 application instances about about a hundred applications So for oh Sorry, my mic is not on. Sorry. Can you hear me now? Okay, perfect. So Cloud Foundry alone is about 600 application instances in prod about 2000 and deaf and it's from the business point of view about a hundred applications So when you think about it, you know for for a large historical insurance company that was really a huge thing to to not only Move into this new technology so quickly but to also be leveraging cloud technology Trying to get this tanker Moving and making the organization more agile really Many times felt like we were trying to make the impossible possible, but we succeeded so That's kind of the the framework But then I'd like to talk to you also a little bit about the people factor and and really what it meant and what what the major challenges were that we encountered in that space and You know in my opinion, it's really a paradigm shift Within our organization that we're facing because we need just a different mindset we need people who Are willing to try new things who especially an insurance company? We don't we don't like taking risk, right? And so to try new things to to make mistakes to learn from those mistakes Those were some of the major challenges and I'll get into a little bit more detail Just in a couple of slides from now So with all of this new stuff and this digital transformation, you know, it felt a little bit like it might have felt For the Beatles when not everybody was on board not not everybody Believed that this was really going to be You know an exciting successful thing So as it comes to people Our experience was that it really is a mindset Thing, you know, it takes a new way of thinking whereas in the past You know people were told what to do how to do it and now it was really more about every single person taking Responsibility and for us to begin to define the framework and the and the sort of the the boundaries within which They could actually do that and do that safely And so what we did was we created We ran a bunch of workshops where we defined roles and responsibilities where we talked about Our mission what you know what you saw in that second slide, not just the new team name but also a logo and a mission that went with it So they could identify themselves with it and we actually went from eight internal employees who had been running our portal environment to 21 Cloud Ops team members. We took all of the strategic initiatives that had been in the hands of external consultants and service providers and they are now all run within that team of 21 folks So that also means we have some legacy Some legacy responsibilities that they still have to cover but they're mostly now focused on Running this platform as a service and continuing to evolve it and automate it and improve it and It's also about in a way of working of course. So we had to in the past We ran long projects, you know never-ending projects that by the by the time we're finally done Who knows what was at the end of it if it was what we really needed or what the customer needed? And now we have the team set up in different clusters so for the different platform topics it's We have a we have a cluster that is Solely focused on all of our legacy core components and we have a product owner that is in charge of running that and then We also have the new the new cloud team Divided up into three clusters the monitoring cluster the runtime cluster and and then also The integration layer So within those four different smaller teams, they are now focused on their respective areas of responsibility and the product owner makes sure that all of the important things that need to be handled are put into the backlog and And prioritized accordingly and the team then all of the engineers help of course work through that backlog and Of course, we need in the way of leadership, right? It's no longer it doesn't work anymore to have somebody there who has to know everything the world has become way too complex To to try and have one person tell the entire team what and how to do things So certainly more of a servant type leadership where the where the whole power Pyramid is turned upside down and it's now about really truly enabling and empowering the team and Taking the roadblocks out of the way and just letting them do things and it's really truly amazing. I mean, I've watched these guys Put this environment together. Yeah, we had a lot of external expertise that we brought in at the very beginning but the the team members Both the internal as well as the external folks work together so well and created this incredible Platform for us and continue to Improve and evolve it. It's it's been just incredible to to watch them do that and have fun with it, you know So one of the really important pieces as I mentioned before is to to make sure especially when you have a culture Which I think most companies do where you know, it's it's not really cool to make mistakes to To make sure there's no blaming that we Take those mistakes and we make sure we learn from them and We also have found that we have had to break down the silos, you know it's It doesn't work anymore to have the ops folks and the and the deaf folks and the architects and the security people all just work in their own little in their own little world, but it's just been tremendously helpful to have them interact on a regular basis and One of the things actually that we've learned one of our mistakes was that in the beginning We were so busy setting up this platform While the training centers were getting staffed and and teams started developing new apps that we did not interact with them enough so a lot of the stuff that could have been avoided You know things that they had to redo because they didn't know how to do it properly Could have been avoided if we had had more of a present from the beginning and we're actually part of those teams or or At least circling through those teams having a present in those training centers Or we just couldn't do it all at the same time, you know so that was definitely a big lesson learned and and and and the other Lesson certainly was to try and figure out how to deal with Legacy team on on one side and the new Team moving into the new technology on the other side and initially the thought was to create two different teams Bless you, but then we realized, you know, if we create two different teams, it's going to be impossible for those folks who are Today in charge of the legacy stuff for them to move into the new world at some point and to actually help move The legacy world into the new world and for everybody to begin to get an end-to-end Understanding of the complexity of our environment. So we we decided to create one really large team Which you could argue that's not really the two pizza team that Amazon always talks about right but for for this first phase We decided to just cluster the way I mentioned earlier and have them focus on on their respective areas of responsibility but to keep them as one team so that they could Interact more and and take along, you know, the legacy folks for the ride and You know initially we also thought that we define this one time and then we know how to do it And then we just go and and that's the end of that right but we realized as as things changed and evolved and the environment grew We had to adjust things and you create those clusters and And create the different roles and have the product owners in charge of the of the various topics and Define the roles of the of the platform engineer and and define what it actually meant now to be a platform engineer and how the product owners would then work together and exchange Their priorities so that they all knew what what each other were working on so it's an evolving It's been an evolving process for us and not necessarily, you know, you define it once and you're done That was certainly a lesson learned And one of the things that we've just recently said was that Not having having moved from the project team now to more of a product team type setup We realized that now that the project is closed We don't have enough of a security presence anymore. So we were just doing a couple of things the other day. We as part of a dojo with pivotal That we implemented and didn't realize we probably should have should have asked security for sign off and We just hadn't thought about it, you know, and and the problem with security folks is that they're so Needed and wanted everywhere that they're not Reachable that they're not around enough and timely enough so one of the things we've decided to do is to create a You know a specific security focus within our team and have one or two folks really focus on that and And represent that security function within within the team. So, yeah, and this is the team You can see on the slide here, you know, some of what I already mentioned that we moved from Project manager to these new roles of product on our platform engineer and so forth and And that's actually been been working really well And here is that here the clusters again that I mentioned earlier With the different products that are associated with that and you can see the monitoring Cluster spans both worlds that on the left the legacy world and on the right the the new world So that's our way of really beginning to pull all of that legacy knowledge Into into the new world So for now, they're focused on their clusters But midterm Goal is to get everybody to understand At least at a superficial level the entire platform as a service Thank you So, can you hear me? That's all right Okay, so I really have to hurry a little now as real No worries 20 minutes already passed, but I think it's it's okay because in our opinion people is Obviously the most important part of of this transformation Undertaking so I'm coming to processes We decided to move from project management to product management and we what we mean by that or what we mean by treating your platform as a product is That you that you handle it differently, so when we implemented the cloud platform in 2016 we followed or we obviously undertook a huge project what was really good because it was a complex unique and new undertaking it has fixed resources and a timeline and now we really Want to to have dynamic goals and decide ourselves what we spend our resources on so therefore we switch to the whole product management Yeah concept and even though we are still a way to go it's working out fine so far for us We also Adopted and experimented with a lot of agile methods, and I think this goes hand-in-hand with all those cloud native Yeah Movement so to say most important method I want to highlight is this pivotal dojo so We not only send application teams to London to pivotal labs, but also got some great engineers to Munich to our office and Had a dojo like an joint engagement over a couple of weeks where we work together and yeah experiment with all of those Agile methods This is only an excerpt of what we actually Are doing right now and experimenting with Also to highlight pair programming and co-location. We really believe this is the way to go not only in terms of quality of what you're building but also in terms of knowledge transfer and Yeah, and I think fun and and efficient work Ritos user interviews you you know all those techniques But we really did this for the like first time in an honest and and serious way So we went to our application developers to their offices Which which are on a different location and interviewed them just listened for for hours And even if they complained we just listened wrote everything down and we condensed the Results and feed it feed it everything into our backlog or forwarded it to different people who were in charge Also to highlight cross-functional teams as Andrea said We try and really try to to bring everyone together IT security is really important We really want to get the officers to be more more of an engineer Basically to be there at the beginning of the process and develop together with us not only saying yes or no at the end Yes or no this way Yeah, and I could talk about all those methods all day long also important and This might seem very basic, but it's actually we believe it's very very important You have to have a beneficial working environment This is our new environment and we are moving to this this work This office in about six weeks. I think really excited about it because this is following the approach of Our agile training centers the new offices where all the application developers are located We have nice bright furniture. We have paring stations We have walls we can write on big screens. We can put our Prometheus Grafana monitoring stuff on Yeah, and we're really happy that we could convince Top management to build this office for the cloud team Because as you can imagine this also costs money But we really believe this this is like a prerequisite for for paring and and have a well-functioning team basically This slide is also very important Because with our new platform we Basically implemented a new working model. There is now a sharp responsibility between the application developers and the cloud ops team as you can see here This is like this layered stack cloud ops the 21 people Are in charge of the virtual infrastructure which we host at Amazon web services right now Our container platforms and we have both cloud foundry and OpenShift and we implemented both in 2016 at the same time And also the platform services around it the additional services like monitoring for example We provide application level monitoring as a service to our customers Or CI CD we have a standardized Pre-provision CI CD pipeline for every team who is joining our platform and to do all of this with 21 people for About 600 developers on the platform right now and the process have to scale and So we really have to to educate people not to route every ticket to us because we're not an ops unit We are cloud ops and we are in charge of the cloud platform and infrastructure But not of your application and this conversation. Yeah, sometimes is kind of intense but We are doing fine so far and everyone is is getting along Also last slide for the processes processes chapter We of course had to develop a lot of new operational processes about our platform So oh, this isn't this isn't all slide actually I updated it But maybe to not confuse you. I'll just put it away So basically we we implemented processes around the platform like technical processes How do we want to do patching scaling updating disaster recovery backup and all of that? We adapted our ITSM processes, which is very important because if you want to have CI CD and everything you cannot Have cap meetings every seven days and force everyone to wait for days to deploy and Also, we yeah, we had to to get the operational processes about the platform Yeah Implemented like onboarding off-boarding support processes for the application developers and we implemented monitoring for all of those and Yeah, it's working fine so far and the platform is actually really stable What we have to improve is the onboarding process. I'm honest because this is mainly a manual process right now, but this was I will put a lot of effort in in building all those processes surrounded So the technology chapter to sum it up is really short because we can see here obviously most important part is a setting up Cloud Foundry or container Technologies we set it up on elastic interest structure and I really just want to Say one sentence about this slide because it's important to us that this new platform is secure because it's a new platform It's in the clouds So we really want to make sure it's it's secure and there's a talk about this. I had last year It's in German unfortunately, so I don't know if if you guys Want to check it out, but it's it's on YouTube and if you want to know more about it Feel free to reach out to me after the talk Last slide for my part More than technologies We see those technologies and this is actually Yeah quite the technologies we implemented in the last 18 months with an alliance You see there are technologies for container container technologies cloud elastic infrastructure CICD monitoring API gateways and what have you even operating systems and we brought everything in We brought a lot of technologies in but I think we found the balance between standardization and individuals individual solutions And also to mention with those technologies being just an enabler we We are following the Rationals or best practices on the right side even if we're not perfect in it We really try to do everything which you can see on the right side So hand it over to Andrea to sum it up okay So with all of that lessons learned and success factors from a Technology perspective as Ludwig just said We're providing a platform for cloud native apps We provide a standardized CICD pipeline fully automated provide additional services and we also follow those best practices as we just heard and The important piece here is that with technology just like the processes and the and the people puzzle It's We're not done. You know we we decided to just Get started do the best we can with the first iteration and we continue to Improve and automate and optimize and every single person who's on that team now That's what their main focus is is to optimize and to automate and to make not only the technology But also the processes better and our external folks that have been Along with us on this journey were of course we had to bring in some expertise because with all this new technology and only a Handful of people on the original team. We couldn't do it all ourselves But for this year one of the main goals was to take that knowledge from from the external folks And really make sure that the knowledge transfer happens within the first six months of this year So that the internal the new team Are actually up and running and can support and manage this platform on their own As far as the processes. Yeah, we now are focused on product versus project management We introduce agile methods. We have Assistant still with that need some coaching on that scrum master roles product owner roles All of that is new especially given that we are operating within the larger organization and the people around us are also Just beginning to learn Pivotal helped out with different projects dojos and You know just the work environment Is definitely something that that's going to make a difference once we can move into the new Workspace and it's really also important for every single one of the team members to be out there to be positively promoting all of this to be an evangelist and to To always, you know create positive interactions with the other folks As far as the people new way of thinking new way of working new way of leadership no blaming There to make mistakes and learn from them and then another really Important piece especially when it came to the recruiting. I think it's a similar situation here in this country in Germany It's been really challenging to find the right people because everybody's looking for them right now So what has been really helpful for us first of all to go through all the different channels Whether be headhunters or whatever but also to get the word out there especially as a as an insurance company You know people who are into this kind of stuff probably don't think about insurance company That's not on the top of their list of cool places to go work Right so to make sure that word gets out about what it is We're doing and how many cool things we're doing within the company has definitely helped To find you know a great mix of people a very diverse team with different backgrounds different skill sets, but even though it's the ops team Which is where we started this journey still have them focus on on moving toward more of a development type mindset And inside the company we can't communicate enough it always feels like you know We need to do more of that, but that's the case with with any change and to just keep going And to keep pushing it and that's one of my favorite slides So in order to be successful with this whole endeavor In in my experience It is absolutely critical that you have the top-down Support otherwise you do nothing but try to create business case papers, right? So you need your top management support in this you have to also get From the bottom up all of your folks on board explain why you're doing it why it's important why it's actually really an awesome and fun endeavor and It's not just an it initiative This really takes the entire organization. It takes the business. It takes it. It takes HR It takes everybody to work together on this transformation of your of your business and you really need a business vision That tells you where you're headed in terms of digital transformation to make this journey successful in our opinion And that's we're hiring if any questions Any questions You know what it really was an hour and we were really lucky in our situation. They just let us do it They provided the budget and They just let us do it. They didn't say how to do it. They just said just go do it and You know, obviously we communicated with them on a regular basis and they were very Present when we needed for them to be present And in terms of also communicating into the business or whatever. So it was really our CIO who was just The one who kicked this off and and and enabled it to happen and got you know, the support of his peers So that we could focus on what we needed to focus on which is you know Get this platform up and running and the team on board and everything, you know, we just told you about Yeah, you don't want them, you know Too too close but supporting Yeah Ideally, yes. Yeah, but we had a bunch of other stuff going on at the same time and just had to Juggle that but that was certainly one of the main Main priorities Actually we talked about that as an entire company. So what I mentioned at the very end, you know To have a vision of a business vision of what a customer centric Digital transformation could look like we all actually worked on defining that together You went top down bottom up, you know, and and we created and then we were also given a what we call a renewal agenda with different Areas of focus to ensure that everybody understood what the what the vision was in the strategy and where we were headed Yeah, any other questions? Feel free to reach out if you have specific. Yeah So if I heard you right you're wondering if if it was hard to get the buy-in from the developers to do the poor pair programming So, you know the the first couple of teams that went to Pivotal As far as I know they volunteered. Oh, yeah, so they volunteered to go and then when they came back, you know, they were so excited and You know that there were enough other people who wanted to Participate and so those were the first ones to move into those training center spaces Now it'll be interesting to see how the whole thing will scale up and expand You know, I know a few other companies who have been pursuing this for a longer time and There's certainly still a lot of work to do in that space particularly. Yeah, so thank you very much for oh, okay I was just going to go through We do get I think one of the things that helps personally. This is not so much a question as you know We oftentimes the questioner behind it It's usually have these big the eyes, you know, we're gonna have fewer defects or a quicker time to resolve the effects Or quicker time to get features in the production of ways of building but I think it's also useful to be really set the tone with Developers like what's the benefit to you other than management said do this otherwise, you know My experience is you know, that's just a recipe for people's certain process You know the way and I get this question Hey, we can get faster video during the airport. I mean, and I was come back with you But What are the trade-offs you where are you today and what are the things that are blocking you as a development team? Is it you know? You're waiting over you know, like the only guys who are keeping code reviews, you know, and then you're having to go back It's defects and I find it useful personally that you kind of say go Explain like here are the things that you're suffering from that you complain about today and here's how this might affect that Let's come back and you know a month or two after Pairing and see how did this you know, how is your experience? Because there has to be something in it for everybody Yeah, oh absolutely absolutely so you know for us it was Both of those things one to have a few people who were really excited about it and came back and talked about it but then you also obviously yeah, you have to continue to explain the why and and Make sure that's clear. Yeah But also definitely helps is to to have pairing stations So all the teams who are on our platform are going to the ATC. They are pairing stations They can choose wherever they want to sit, but there are two screens two keyboards two mice That helps So, thank you all if you have any other questions feel free to talk to us Afterwards and enjoy the rest of your day. Thank you. Thank you