>> Live from Las Vegas; it's the Cube. Covering the AWS Accenture Executive Summit. Brought to you by Accenture. >> Welcome back everyone to the Cube's live coverage of the AWS executive summit here at the Venetian in Las Vegas Nevada. I'm your host Rebecca Knight. We have two guests for this segment. We have Hitoshi Ienaka the CEO of ARISE Analytics and Takuya Kudo the Chief Sciences Officer at ARISE. Thank you both so much for coming on the program. >> Thank you. >> So I want to start by having you tell our viewers a little bit more about ARISE analytics. >> Well ARISE analytics is a joint venture between KDDI and Accenture. Well last, well last year we established a company yeah. That's family. >> Right and that's you know kind of we provide like tying the capabilities and the KDDI is kind of number two mobile network operator in Japan, has 50 million subscribers, massive data. So that's there a lot of room to cook but they don't have enough capability to support that. So that's why we kind of married together. >> And it helps companies leverage a wealth of knowledge resources and data between firms to bring about digital transformation. >> Right. >> That's what you're doing. So talk a little bit about what you've seen so far. >> Well so we have two assets, KDDI has, well big data and well Accenture has, well a lot of analytic skills. So using this well these assets, we built our integrated analytics platform hosted on a eda-brais. And what our first challenge was to deduce, channel out to the other operators and were which caused a challenge risk to well more than 40 million subscribers and by digging into that data and using machine learning origin and our data includes (mumbles) and life style service usage. And well, we optimize customer channels and contact timing and well to target customers efficiently. And well we well we tried art of well, other event well art of >> (mumbles) >> Yeah yeah. >> Yeah (mumbles) marketing. >> Okay. >> Yeah and we can get a good result and well it was not only due to our activities but only last year, only KDDI well could increase the market share among three network operators in Japan. That is our our achievements yeah. >> That's very impressive! So can you talk a little bit about the initial pilot in particular what you saw. Taku, do want to? >> Right so like as he mentioned like we have two work stream gigantic work stream. One is for consumer facing right. So customer chai and the you know out of on three marketing's or like recommendation engines based upon this stream data because we have massive like this is a consumption data too. Not just about like you know one handset data. In another work stream is a B2B, a business domain which is sounds like not related to mobile network operators but they have massive network to sell to B2B customer. So we utilize those gigantic data, combine those maybe I can mention but data but combine those data creating new service model. So that's quite a new IOT initiatives for B2B layers and consumer initiatives you know to support ongoing current business. >> And you're using this in a variety of sectors in particular I wanted you to ask you about one that you're doing with Toyota and a taxi service. >> Right so (mumbles) so yeah that that one is like five years like example because a, unless otherwise, I don't think that new business model to compete with Uber never happened right? So KDDI provide like Maura Handu said like location data over like you indigenous subscribers creating some, you know demand side riders for (mumbles) right? Over there, on top of that Toyota's transact log, which is technically like kinematics data provide like supply side which is cause, right? Focusing model and taxi also provide like meters, where customer riders get in and get off and combine those three completely different cable and data sets. >> But also with things like weather and those kinds of other >> Exactly yeah. >> outside. >> Open data too. And combine those data sets. We in, we provide, Accenture provides like talents and creating completely new forecasting model it's called AI taxi dispatch model. So now if you go to Tokyo, majority of taxi has our algorithm like Arizona takes in, you know KDDI and Accenture provide it. >> So that's very cool! Can you talk a little bit about what you've learned, about, in terms of when the weather is like this, taxis happen this >> Yeah, so it's of course weather has massive impact over, like if it's mornings specifically lane, it boosts like demands and also events. We have also events data. Maybe I don't know concerts, some famous singer, celebrities came and it's you know boost like riders demands. So that's actually significant impact of our demand focused model. Rather than using pushing like Uber, you strike you know app, mobile app. we actually treated as (mumbles) like taxi actually go because taxi driver and I can see where is a hot spot to pick up riders. And that's what we try to do. So based up on those, you know people don't even have like maybe like my father's age right, that don't have a smartphone they can get the benefit universality right. So that's the base concepts to provide Universal model to those you know without these >> So even people lacking technology >> Right exactly. >> Can still reap the benefits of this kind of approach. >> (mumbles) is universality so that's also our business strategy. Yeah. >> So you're also using this approach in a manufacturing environment. >> Yeah that's right. We are also working with some manufacturing factory. On the factory field were experienced workers can detect machine breakdown before they occur. But well how can that not be passed on to less experienced employees? So we created a live predictive maintenance which alerts companies ahead of time to pre potential breakdowns. Sensors (mumbles) about things like vibrations, temperature and electrical current. The collected data is analyzed by the AI system. So in this way the prediction of machine (mumbles) can be performed by almost anyone. Well it used to be others by only experienced employees before, yeah. >> So it not only helps the company know when a machine is going to fail, it also empowers the employee to fix it him or herself. >> Right it's a preventive way and so it's up and running over the ad-abreis. We use kinesis in late shift you know, learn the functions and over EC2. So that's completely free stock over ad-abreis capability too. >> So what you're describing sounds like it requires a lot of collaboration, a lot of deep relationship building between not only Accenture and KDDI but also the clients that you're working with. Can you describe how you all work together? >> Right. So maybe I'm going to provide that information. So like of course like KDDI's employee has specific domain knowledge and we provide like you know like data science capabilities and also like maybe through the interview right, found workers or like taxi, they have specific domain knowledge So combine those collaboration. It's called two in the box and we collaboratively paired each employees and you know supply the knowledge each other so that's it. Just one is not enough but as a team integrated over database and created a very strong team and that's a you know we try to cherish and that's culture. And the two boost the data science, data driven companies decision-making process. >> So i think our viewers are pretty amazed and impressed with what's going on. But in this era of 5G and IOT, what's next, what are you working at? It's a relatively new partnership. What are what are some of the most exciting things in the pipeline? >> So the (laughs) the very strategic so we strategizing right now in terms of 5G in IOT. But definitely one of the pieces could be like deep learning right? And also about your realities which nobody has done before. So that's where we try to collaborate with other sectors, industries, to create a new. And to do so we need a massive like computation power like GPU servers and we have to rely on the ad-abreis because otherwise we cannot achieve those goals and specifically 5G maybe changing in the game. Maybe like you know low latency and you know wireless connectivity, you know we don't need connections so maybe the factory lining assembly lines. You know completely change the way crispy like edge computing no more. Maybe like for computing, right, in between like Saba and edge because of the 5G. I don't know but we are strategizing now in a very exciting moment. We are doing right now. >> Indeed it is. >> Yeah. >> Well Hitoshi, Taku, thank you so much for coming on the Cube. This was a lot of fun. >> Thank you very much. I'm Rebecca Knight. Stay tuned for more of the Cube's live coverage of the AWS Executive Summit coming up just after this. (Uptempo music)