hey welcome to analytics unleashed i'm robert christensen your host today thank you for joining us today we have three quick wins that drive big gains in the enterprise workloads and today we have olaf with erickson we have john with orok and we have dragon with dxc welcome thank you for joining me gentlemen yeah good to be here thank you thank you good to have you hey olaf let's start off with you what big problems are you trying to solve today that are doing for those quick wins what are you trying to do today top top of mind yeah when we started looking into this microservices for our financial platform we immediately saw the challenges that we have and we wanted to have a strong partner and we have a good relationship with hp before so we turned to hp because we know that they have the technical support that we need the possibilities that we need in our platform to fulfill our requirements and also the reliability that we would need so tell me i think this is really important you guys are starting into a digital wallet space that correct yeah that's correct so we are in a financial platform so we are spanning across the world and delivering our financial services to our end customers well that's not classically what you hear about ericsson diving into what's really started you guys down that path and specifically these big wins around this digitization no what what we could see earlier was that we have a mobile networks right so we have a lot of a strong user base within them uh both kind of networks and in the where we started in the emerging markets uh you normally they have a lot of unbanked people and that people also were the ones that you want to target so be able to instead of going down and use your cash for example to buy your fruits or your electricity bill etc you could use your mobile wallet and and that's how it all started and now we're also turning into the emerged markets also like the western side part of worlds etc that's fantastic and i hey i want to talk to john here john's with o'rock and he's the one of those early adopters of those container platforms for the uh in the united states here the federal government tell us a little bit about that program and what's going on with that john yeah sure absolutely appreciate it yeah so with orock what we've done is we developed one of the first fedramp authorized container platforms that runs in our moderate and soon to be high cloud and what that does is building on the israel platform gave us the capability of offering customers both commercial as well as federal the capability and the flexibility of running their workloads in a you know as a service model where they can customize and typically what customers have to do is they have to either build it internally or if they go to the cloud they have to be able to take what resources are available then tweak to those designs to make what they need so in this architecture built on open source and with our own infrastructure we offer you know very low cost zero egress capability but the also the workload processing that they would need to run data analytics machine language and other types of high performance processing that typically they would need as we move forward in this computer age so john you you touched on a topic that's i think is really critical and you had mentioned open source why is open source a key aspect for this transformation that we're seeing coming up in like the next decade yeah sure yeah with open source we shifted early on to the company to move to open source only to offer the flexibility we didn't want to be set on one particular platform to operate within so we took and built the cloud infrastructure we went with open source as an open architecture that we can scale and grow within because of that we were one of the very first fedramp authorizations built on open source not on a specific platform and what we've seen from that is the increased performance capability that we would get as well as the flexibility to add additional components that typically you don't get on other platforms so it was a it was a good move we went with and one that the customer will definitely benefit from that that's that's huge actually because performance leads to better cost and better cost leads better performance around that i i'm just super super happy with all the advanced work that you always are doing there is fantastic and dragon so so you're in a space that i think is really interesting you're dealing with what everybody likes to talk about that's autonomous vehicles you're working with automobile manufacturers you're dealing with data at a scale that is unprecedented can you just open that door for us to talk to about these big big wins that you're trying to get over the line with these enterprises yeah absolutely and um thank you robert we approach uh leveraging esmeral from the data fabric angle we practically have a fully integrated the esmeral data fabric into our robotic drive solution rewarding drive solution is actually a game changer as you've mentioned in accelerating the development of autonomous driving vehicles it's a an end-to-end hyper-scale machine learning and ai platform as i mentioned based on the esmeralda data fabric which is used by the some of the largest manufacturers in the world for development of their autonomous driving algorithms and i think we all in technology i think and following up at the same type of news and research right across the globe in in this area so we're pretty proud that we're one of the leaders in actually providing uh hyperscale machine learning platforms for uh kind manufacturers some of them i cannot talk about but bmw is one of uh one of the current manufacturers that we provide uh these type of solutions and they have publicly spoken about their uh d3 platform uh data driven development platform uh just to give you an idea um of the scale as robert mentioned uh daily we collect over 1.5 petabytes of data of raw data did you say daily data daily the storage capacity is over 250 petabytes and growing uh there's over 100 000 cores and over 200 gpus in the in in the compute area um over 50 50 petabytes of data is delivered every two weeks into a hardware in loop right for testing and we have daily uh thousands of engineers and data scientists accessing the relevant data and developing machine learning models on the daily basis right part of it is the simulation right simulation cuts the cost as well as the uh time right for developing of the autonomous uh driving algorithms and uh the the simulations are taking probably 75 percent of the research uh that's being done on this platform that's amazing dragon i i i i the more i get involved with that and i've been part of these conversations with a number of the folks that are involved with it i i computer science me my geekiness my little propeller head starts coming out i might just blows my mind and i think so i'm going to pivot back over to olaf oh left so you're talking about something that is a global network of financial services okay correct and the flow of transactional typically non-relational transactional data flows to actual transactions going through you have issues of potential fraud you have issues a safety and you have multi-geographic regional problems with data and data privacy how are you guys addressing that today so so to answer that question today we have managed to solve that using the container platform to together with the data fabric but as you say we need to span across different regions we need to have the data as secure as possible because we have a lot of legal aspects to look into because if our data disappears but your money is also disappearing so it's a really important area for us with the security and the reliability of the platforms so so that's why we also went this way to make sure that we have this strong partner that could help us with this because just looking at where we are deployed in in more than 23 countries today and and we it's processing more than 900 million us dollars per day in our systems currently so it is a lot of money passing through and you need to take security in a it's as it's a very important point right it really is it really is and so uh john i mean you you uh obviously are dealing with you know a lot of folks that have three letters as acronyms around the government agencies and uh they range in various degrees of certa of security when you say fedramp i mean what could you just uh articulate why the esmerald platform was something that you selected to go to that fedrak compliant container platform because i think that's that that kind of speaks to the to the industrial strength of what we're talking about yeah it all comes down to being able to offer a product that's secure that the customers can trust and when we went with fedramp fedramp has very stringent security requirements that have monthly poems which are performance reviews and and updates that need to be done if not on a daily basis on a monthly basis so the customers there's a lot that goes on behind the scenes that they don't are able to articulate and what by selecting the hp esmerald platform for containers um one of the key strengths that we looked at was the esmo fabric and it's all about the data it's all about securing the data moving the data transferring the data and from a customer's perspective they want to be able to operate in an environment that they can trust no different than being able to turn on their lights or making sure there's water in their utilities you know containers with the israel platform built on orok's infrastructure gives that capability fedramp enables the security tied to the platform that we're able to follow so it's government uh guided which includes this and many and over hundreds of controls that typically you know the customers don't have time or the capability to address so our commercial customers benefit our federal customers you know that you discuss they're able to follow and check the box to meet those requirements and the container platform gives us a capability where now we're able to move files which we'll hear about through the optimal fabric and then we're able to run the workloads in the containers themselves and give isolation and the security element of fed wrapping esmeral gave us that capability in order to paint that environment fedramp authorized that the customers benefit from from security so they have confidence in running their workloads using their data and able to focus on their core job at hand and not worry about their infrastructure the fundamental requirement isn't it that that isolation between that compute and storage and going up a layer there in in a way that provides them a set of services that they can i wouldn't say set it and forget it but really had the confidence that what they're getting is the best performance for the dollars that they're spending uh john my hat's off to what the work that you all do in there thank you we appreciate it yeah yeah and dragon i want to i wanted to pivot a little bit here because you are primarily the the operator what i consider one of the largest data fabrics on the on the planet for that matter um and i just want to talk a little bit about the openness of our architecture right of all the multiple protocols that we support that allow for you know you know some people may have selected a different set of application deployment models and virtualization models that allow to plug into the data fabric you know it did can you talk a little bit about that yeah and i i think um in my mind right um to operate uh such a uh data fabric at scale right um there were three key elements that we were looking for right uh that we found in uh esmeralda fabric ring the first one was a speed cost and scalability right the second one was the globally distributed data lake or ability to distribute data globally and third was certainly the strength of our partnership with with hpe in this case right so if you look at the uh as well data fabric it's it's fast it's cost effective and it's certainly highly scalable because we as you just mentioned stretch the uh sort of the capabilities of the data fabric to hundreds of petabytes and over a million the data points if you will and it important what was important for us was that the esmeralda fabric actually eliminates the need for multiple vendor solutions which would be otherwise required right because it provides integrated file system database or or a data lake right and the data management on top of it right usually you would probably need to incorporate multiple tools right from different vendors and the file system itself it's it's so important right when you're working at scale like this right and honestly in our research maybe there are three file systems in the world that can support uh this kind of size of the auto data fabric the distributed data lake was also important to us and the reason for that is you can imagine that these large car manufacturers are testing and have testing vehicles all around the world right they're not just doing it locally around the uh their data their id centers right so uh collecting the data and this 1.5 petabytes example right uh for for bmw on a daily basis it's it's it's really challenging unless you have the ability to actually leverage the data in a distributed data like fashion right so data can basically reside in different data centers globally or even on-premise and in cloud environments which became uh very important later because a lot of this car manufacturers actually have oems right that would like to get either portions of the data or get access to the data in a in different environments not necessarily in their data center um and truly i think uh to build something at this scale right uh you you need a strong partner and we certainly had that in hpe and uh we got the comprehensive support right for uh for the software um but but more importantly i think uh partner that clearly understood uh criticality of the data fabric trend and the need for the vice fast response right to our clients and you know jointly i think we met all the challenges and it's so doing i think we made the esmo data fabric a much better and stronger product over the over the last few years that's fantastic thank you dragon appreciate it uh hey so if we're going to wrap up here any last words olaf do you want to share with us no looking forward now in from our perspective on helping out with the kobe 19 situation that we have uh enabling people to still be in the market without actually touching each other and and and leaving maybe for action market and being at home etc doing those transactions that's great thank you john in last comment yeah thanks yeah uh look for uh a joint offering announcement coming up between hpe and orok where we're going to be offering sandbox as a service where the data analytics and machine language where people can actually test drive the actual environment as a service and if they like it then they can move into a production-wise environment so stay tuned for that that's great john thank you for that and hey dragon last words yeah last words um we're pretty happy what we have done already for car manufacturers we're taking this solution right in terms of the uh distributed data-like capabilities as well as the uh hyperscale machine learning and ai platform to other industries and we hope to do it jointly with you well we hope that you do it with us as well so thank you very much everybody gentlemen thank you so much for joining us i appreciate it thank you very much thank you very much hey this is robert christensen with analytics unleashed i want to thank all of our guests here today and we'll catch you next time thank you for joining us bye [Music] [Music] [Music] easy [Music] you