>> live from San Francisco, California. It's the Q covering the IBM Chief Data Officer Summit brought to you by IBM. >> Welcome back to Fisherman's Fisherman's Wharf in San Francisco. Everybody, my name is David wanted. You're watching the Cube, the leader in live tech coverage, you ought to events. We extract the signal from the noise. We're here. The IBM CDO event. This is the 10th anniversary of this event. Caitlin Hallford is here. She's the director of a I Accelerator and client success at IBM. Caitlin, great to see you again. Wow. 10 years. Amazing. They and Carlo Apple Apple Glace e is here. Who is the program director for data and a I at IBM. Because you again, my friend. Thanks for coming on to Cuba. Lums. Wow, this is 10 years, and I think the Cube is covered. Probably eight of these now. Yeah, kind of. We bounce between San Francisco and Boston to great places for CEOs. Good places to have intimate events, but and you're taking it global. I understand. Congratulations. Congratulations on the promotion. Thank you. Going. Thank you so much. >> So we, as you know well are well, no. We started our chief date officer summits in San Francisco here, and it's gone 2014. So this is our 10th 1 We do two a year. We found we really have a unique cohort of clients. The join us about 100 40 in San Francisco on the spring 140 in Boston in the fall, and we're here celebrating the 10th 10 Summit. >> So, Carlo, talk about your role and then let's get into how you guys, you know, work together. How you hand the baton way we'll get to the client piece. >> So I lead the Data Center League team, which is a group within our product development, working side by side with clients really to understand their needs as well developed, use cases on our platform and tools and make sure we are able to deliver on those. And then we work closely with the CDO team, the global CEO team on best practices, what patterns they're seeing from an architecture perspective. Make sure that our platforms really incorporating that stuff. >> And if I recall the data science that lead team is its presales correct and could >> be posted that it could, it really depends on the client, so it could be prior to them buying software or after they bought the software. If they need the help, we can also come in. >> Okay, so? So it can be a for pay service. Is that correct or Yeah, we can >> before pay. Or sometimes we do it based on just our relation with >> It's kind of a mixed then. Right? Okay, so you're learning the client's learning, so they're obviously good, good customers. And so you want to treat him right >> now? How do you guys work >> together? Maybe Caitlin, you can explain. The two organizations >> were often the early testers, early adopters of some of the capabilities. And so what we'll do is we'll test will literally will prove it out of skill internally using IBM itself as an example. And then, as we build out the capability, work with Carlo and his team to really drive that in a product and drive that into market, and we share a lot of client relationships where CEOs come to us, they're want advice and counsel on best practices across the organization. And they're looking for latest applications to deploy deploy known environments and so we can capture a lot of that feedback in some of the market user testing proved that out. Using IBM is an example and then work with you to really commercialized and bring it to market in the most efficient manner. >> You were talking this morning. You had a picture up of the first CDO event. No Internet, no wife in the basement. I love it. So how is this evolved from a theme standpoint? What do you What are the patterns? Sure. So when >> we started this, it was really a response. Thio primarily financial service is sector regulatory requirements, trying to get data right to meet those regulatory compliance initiatives. Defensive posture certainly weren't driving transformation within their enterprises. And what I've seen is a couple of those core elements are still key for us or data governance and data management. And some of those security access controls are always going to be important. But we're finding his videos more and more, have expanded scope of responsibilities with the enterprise they're looked at as a leader. They're no longer sitting within a c i o function there either appear or, you know, working in partnership with, and they're driving enterprise wide, you know, initiatives for the for their enterprises and organizations, which has been great to see. >> So we all remember when you know how very and declared data science was gonna be the number one job, and it actually kind of has become. I think I saw somewhere, maybe in Glass door was anointed that the top job, which is >> kind of cool to see. So what are you seeing >> with customers, Carlo? You guys, you have these these blueprints, you're now applying them, accelerating different industries. You mentioned health care this morning. >> What are some >> of those industry accelerators And how is that actually coming to fruition? Yes. >> So some of the things we're seeing is speaking of financial clients way go into a lot of them. We do these one on one engagements, we build them from custom. We co create these engineering solutions, our platform, and we're seeing patterns, patterns around different use cases that are coming up over and over again. And the one thing about data science Aye, aye. It's difficult to develop a solution because everybody's date is different. Everybody's business is different. So what we're trying to do is build these. We can't just build a widget that's going to solve the problem, because then you have to force your data into that, and we're seeing that that doesn't really work. So building a platform for these clients. But these accelerators, which are a set of core code source code notebooks, industry models in terms a CZ wells dashboards that allow them to quickly build out these use cases around a turn or segmentation on dhe. You know some other models we can grab the box provide the models, provide the know how with the source code, as well as a way for them to train them, deploy them and operationalize them in an organization. That's kind of what we're doing. >> You prime the pump >> prime minute pump, we call them there right now, we're doing client in eights for wealth management, and we're doing that, ref SS. And they come right on the box of our cloudpack for data platform. You could quickly click and install button, and in there you'll get the sample data files. You get no books. You get industry terms, your governance capability, as well as deployed dashboards and models. >> So talk more about >> cloudpack for data. What's inside of that brought back the >> data is a collection of micro Service's Andi. It includes a lot of things that we bring to market to help customers with their journey things from like data ingestion collection to all the way Thio, eh? I model development from building your models to deploying them to actually infusing them in your business process with bias detection or integration way have a lot of capability. Part >> of it's actually tooling. It's not just sort of so how to Pdf >> dualism entire platform eso. So the platform itself has everything you need an organization to kind of go from an idea to data ingestion and governance and management all the way to model training, development, deployment into integration into your business process. >> Now Caitlin, in the early days of the CDO, saw CDO emerging in healthcare, financialservices and government. And now it's kind of gone mainstream to the point where we had Mark Clare on who's the head of data neighborhood AstraZeneca. And he said, I'm not taking the CDO title, you know, because I'm all about data enablement and CDO. You know, title has sort of evolved. What have you seen? It's got clearly gone mainstream Yep. What are you seeing? In terms of adoption of that, that role and its impact on organizations, >> So couple of transit has been interesting both domestically and internationally as well. So we're seeing a lot of growth outside of the U. S. So we did our first inaugural summit in Tokyo. In Japan, there's a number of day leaders in Japan that are really eager to jump start their transformation initiatives. Also did our first Dubai summit. Middle East and Africa will be in South Africa next month at another studio summit. And what I'm seeing is outside of North America a lot of activity and interest in creating an enabling studio light capability. Data Leader, Like, um, and some of these guys, I think we're gonna leapfrog ahead. I think they're going to just absolutely jump jump ahead and in parallel, those traditional industries, you know, there's a new federal legislation coming down by year end for most federal agencies to appoint a chief data officer. So, you know, Washington, D. C. Is is hopping right now, we're getting a number of agencies requesting advice and counsel on how to set up the office how to be successful I think there's some great opportunity in those traditional industries and also seeing it, you know, outside the U. S. And cross nontraditional, >> you say >> Jump ahead. You mean jump ahead of where maybe some of the U. S. >> Absolute best? Absolutely. And I'm >> seeing a trend where you know, a lot of CEOs they're moving. They're really closer to the line of business, right? They're moving outside of technology, but they have to be technology savvy. They have a team of engineers and data scientists. So there is really an important role in every organization that I'm seeing for every client I go to. It's a little different, but you're right, it's it's definitely up and coming. Role is very important for especially for digital transformation. >> This is so good. I was gonna say one of the ways they are teens really, partner Well, together, I think is weaken source some of these in terms of enabling that you know, acceleration and leap frog. What are those pain points or use cases in traditional data management space? You know, the metadata. So I think you talk with Steven earlier about how we're doing some automated meditate a generation and really using a i t. O instead of manually having to label and tag that we're able to generate about 85% of our labels internally and drive that into existing product. Carlos using. And our clients are saying, Hey, we're spending, you know, hundreds of millions of dollars and we've got teams of massive teams of people manual work. And so we're able to recognize it, adopts something like that, press internally and then work with you guys >> actually think of every detail developer out there that has to go figure out what this date is. If you have a tool which we're trying to cooperate the platform based on the guidance from the CDO Global CEO team, we can automatically create that metadata are likely ingested and provide into platform so that data scientists can start to get value out >> of it quickly. So we heard Martin Schroeder talked about digital trade and public policy, and he said there were three things free flow of data. Unless it doesn't make sense like personal information prevent data localization mandates, yeah, and then protect algorithms and source code, which is an I P protection thing. So I'm interested in how your customers air Reacting to that framework, I presume the protect the algorithms and source code I p. That's near and dear right? They want to make sure that you're not taking models and then giving it to their competitors. >> Absolutely. And we talk about that every time we go in there and we work on projects. What's the I p? You know, how do we manage this? And you know, what we bring to the table with the accelerators is to help them jump start them right, even though that it's kind of our a p we created, but we give it to them and then what they derive from that when they incorporate their data, which is their i p, and create new models, that is then their i. P. So those air complicated questions and every company is a little different on what they're worried about with that, so but many banks, we give them all the I P to make sure that they're comfortable and especially in financial service is but some other spaces. It's very competitive. And then I was worried about it because it's, ah, known space. A lot of the algorithm for youse are all open source. They're known algorithms, so there's not a lot of problem there. >> It's how you apply them. That's >> exactly right how you apply them in that boundary of what >> is P, What's not. It's kind of >> fuzzy, >> and we encourage our clients a lot of times to drive that for >> the >> organisation, for us, internally, GDP, our readiness, it was occurring to the business unit level functional area. So it was, you know, we weren't where we needed to be in terms of achieving compliance. And we have the CEO office took ownership of that across the business and got it where we needed to be. And so we often encourage our clients to take ownership of something like that and use it as an opportunity to differentiate. >> And I talked about the whole time of clients. Their data is impor onto them. Them training models with that data for some new making new decisions is their unique value. Prop In there, I'd be so so we encourage them to make sure they're aware that don't just tore their data in any can, um, service out there model because they could be giving away their intellectual property, and it's important. Didn't understand that. >> So that's a complicated one. Write the piece and the other two seem to be even tougher. And some regards, like the free flow of data. I could see a lot of governments not wanting the free flow of data, but and the client is in the middle. OK, d'oh. Government is gonna adjudicate. What's that conversation like? The example that he gave was, maybe was interpolate. If it's if it's information about baggage claims, you can you can use the Blockchain and crypt it and then only see the data at the other end. So that was actually, I thought, a good example. Why do you want to restrict that flow of data? But if it's personal information, keep it in country. But how is that conversation going with clients? >> Leo. Those can involve depending on the country, right and where you're at in the industry. >> But some Western countries are strict about that. >> Absolutely. And this is why we've created a platform that allows for data virtualization. We use Cooper nannies and technologies under the covers so that you can manage that in different locations. You could manage it across. Ah, hybrid of data centers or hybrid of public cloud vendors. And it allows you to still have one business application, and you can kind of do some of the separation and even separation of data. So there's there's, there's, there's an approach there, you know. But you gotta do a balance. Balance it. You gotta balance between innovation, digital transformation and how much you wanna, you know, govern so governs important. And then, you know. But for some projects, we may want to just quickly prototype. So there's a balance there, too. >> Well, that data virtualization tech is interesting because it gets the other piece, which was prevent data localization mandates. But if there is a mandate and we know that some countries aren't going to relax that mandate, you have, ah, a technical solution for that >> architecture that will support that. And that's a big investment for us right now. And where we're doing a lot of work in that space. Obviously, with red hat, you saw partnership or acquisition. So that's been >> really Yeah, I heard something about that's important. That's that's that's a big part of Chapter two. Yeah, all right. We'll give you the final world Caitlyn on the spring. I guess it's not spring it. Secondly, this summer, right? CDO event? >> No, it's been agreed. First day. So we kicked off. Today. We've got a full set of client panel's tomorrow. We've got some announcements around our meta data that I mentioned. Risk insights is a really cool offering. We'll be talking more about. We also have cognitive support. This is another one. Our clients that I really wanted to help with some of their support back in systems. So a lot of exciting announcements, new thought leadership coming out. It's been a great event and looking forward to the next next day. >> Well, I love the fact >> that you guys have have tied data science into the sea. Sweet roll. You guys have done a great job, I think, better than anybody in terms of of, of really advocating for the chief data officer. And this is a great event because it's piers talking. Appears a lot of private conversations going on. So congratulations on all the success and continued success worldwide. >> Thank you so much. Thank you, Dave. >> You welcome. Keep it right there, everybody. We'll be back with our next guest. Ready for this short break. We have a panel coming up. This is David. Dante. You're >> watching the Cube from IBM CDO right back.

>> Live from San Francisco, it's theCUBE. Covering IBM Think 2019. Brought to you by IBM. >> Hey everyone, welcome back to the live coverage here in Moscone North in San Francisco for IBM Think. This is theCUBE's coverage. I'm here with Dave Vellante. I've got two great guests here, Carlos Guevara, chief data officer, Claro Columbia, and Carlo Appugliese- Appugliese? >> Appugliese, yeah, good. That's good. >> Engagement Manager, IBM's Data Science Elite Team, customer of IBM, conversation around data science. Welcome to theCube, thanks for joining us. >> Thanks for having us. >> Thank you. >> So we're here the streets are shut down. AI Anywhere is a big theme, Multi-Cloud, but it's all about the data everywhere. People trying to put end-to-end solutions together to solve real business problems. Data's at the heart of all this. Moving data around from cloud to cloud, using AI and technology to get insights out of that. So, take a minute to explain your situation, what you guys are trying to do. >> Okay, okay, perfect. Right now we're working a lot about the business theme, because we need to use the machine learning models or the artificial intelligence, to take best decisions for the company. We were working with Carlo and Sean Muller in order to know how can we divide the customers who leave the company. Because, for us, it's very important, to maintain our customer, to know how their behavior is from them, and their artificial intelligence is an excellent way to do it. We have a lot of challenge about that, because, you know, we have a lot of data, different systems that are running the data, but we need to put all the information together to run the models. The Elite Team that Carlo is leading right now is helping us a lot because, we know how to handle data, we know how to clean the data, we know how to do the right governance for the data and the IBM Equinix is very compromised with us in order to do that. Sofie, that is one of the engineers that is very close to us right now. She was working a lot with my team in order to run the models. Susan, she was doing a lot for our middleware, FITON, and right now we are trained to do it in over the Hadoop system, running the spark, and that is the good way that we are thinking that it's going to get the goal for us We need to maintain our customers. >> So you guys are the largest telecommunications piece, Claro in Mexico for voice and home services-- >> Yeah. >> Is that the segments you guys are targeting? >> Yeah, yeah. >> And the scope size of, how big is that? >> Claro is the largest company in Columbia for telecommunications. We have maybe 50 million customers in Columbia, more than 50% of their market share. Also, where we have many, maybe 2.5 millions of homes in Columbia, that is more than the 50% of the customers for home services. And you know that is a big challenge for us because the competitors are all the time trying to take our customers and the churn, it also adds to us, and how to avoid that and how to do the artificial intelligence to do it, machine learning is a very good way to do that. >> So, classic problem in telecommunications is churn, right, so it's a data problem, so how did it all come about? So these guys came to you and-- >> Yeah, so they came to us, and we got together, we talked about the problem, and churn was at the top, right, these guys have a ton of data. So what we did was, the team got together, we had, really the way the Data Science Elite Team works is we really help clients in three areas. It's all about the right skills, the right people, the right tools, and then the right process. So we put together a team, we put together some Agile approaches, and what we're going to do, and then we started by spinning up an environment, we took some data in, we took their, and there as a lot of data, it as a terabytes of data. We took their user data, we took their users' usage data, which is like how many texts, cellphone, and then billing data, we pulled all that together in an environment, then the data scientists, alongside with Carlos' team, really worked on the problem. And they addressed it with machine learning obviously, targeting churn, they tried a variety of models, but XGBoost ended up being one of the better approaches. And we came up with pretty good accuracy, about 90, 92% precision on the model. >> On predicting-- >> On predicting churn-- >> Yeah, churn, and also, what did you do with that data? >> That is a very good question because, the company is preparing to handle that. I have a funny history, I say to the business people, okay these customers are going to leave the company, and I forget about that, and two months later, I was asking okay, what happened, they say okay, your model is very good, all the customers goes. Oh my God, what is happening with that? They weren't working with information, that is the reason we're thinking that the good ways to think from the right to the left, because which is the purpose, the purpose is to maintain our customers, and in that case we lose 50,000 customers because we didn't do nothing. We are close in the circle, we are taking care about that, prescriptive models to have helped for us to do it. And okay, maybe that is an invoice problem, we need to correct them, to fix the problem, in order to avoid that, but the first part is to predict, to get in a score, for the churn, and to handle that with the people. Obviously, working also, at the root cause analysis, because we need the churn to fix from the root. >> Carlos, what goes through the scope of, like, just the project because, this is a concern we see in the industry, I got a lot of data, how do I attack it, what's the scope? You just come in, ingest it into a data lake, how do you get to the value of these insights quickly, because obviously they are starving for insights, take us through that quick process. >> Well, you know, every problem's a little different, we help hundreds of clients in different ways, but this particular problem, it was a big data problem, we knew we had a lot of data, they had a Hadoop environment, but some of the data wasn't there. So what we did was, is we spun up a separate environment, we pulled some of the big data in there, we also pulled some of the other data together, and we started to do our analysis on that, kind of separately in the cloud, which was a little different, but we're working now to push that down into their Hadoop data lake, because not all the data's there, but some of the data is there, and we want to use some of that computing network to-- >> So you had to almost do an audit on those, figure out what you want to pull in first, >> Absolutely. >> Tie it to the business, on the business side, what were you guys like? Waiting for the answers, or like, what was some of the, on your side of the process, how did it go down? >> Thinking about our business, we were talking a little bit about that, about the architecture to handle that, ICP for Data within that is a very good solution for that, because we need infrastructure to help us, in order to get the answers because finally, we have a question, we have questions, why the customers are leaving us. And, the answer was the data, and the data was handled in a good way, with governance, with data cleaning, with the right models to do that, and right now, our concern is business action, and business offer, because the solution for the company is that we, obviously new products are coming from the data. >> So 10 years ago, you probably didn't have a Hadoop cluster to solve this problem, the data was, maybe it was in a data warehouse, maybe it wasn't, and you probably weren't a chief data officer back then, you know, that role kind of didn't exit. So a lot has changed, in the last 10 years. My question is, do you, first of all, I'd be interested in your comment on that, but then, do you see a point in which you can now take remedial action, or maybe even automate some of that remedial action using machine intelligence and that data cloud, or however else you do it, to actually take action on behalf of the brand, before humans, or without even human involvement, did you foresee the day? >> Yeah, so, just a comment on your thought about the times you know, I've been doing technology for 20 something years, and you know, data science is something that's been around but it's kind of evolved in software development. My thought is, you know, we have these roles of data scientist, but a lot of the feature engineer and data prep does require traditional people that were DBAs and now data engineers, and a variety of skills come together, and that's what try to do in every project. Just to add to that comment. As far as predicting ahead of time, like I think you were trying to say, what data, help me understand your question. >> So you've got 93% accuracy, okay, so, I presume you take that, you give it to the business, business says okay, let's maybe, you know, reach out to them, maybe do a little incentive, or, what kind of action can the machines take action on behalf of your brand, do you foresee a day when that could happen? >> Ah. >> Ah, okay. >> Yeah, so my thought is, for Claro Columbia and Carlos, but obviously this is, to me, remain, is the predictive models we've built will obviously be deployed, and then it would interact with their digital mobile applications, so in real time it'll react for the customers. And then, obviously, you know you want to make sure that Claro and company trust that, and it's making accurate predictions, and that's where a lot more, you know we have to do some model of validation, and evaluation of that, so they can begin to trust those predictions. I think is where we're-- >> Guys. I want to get your thoughts on this because you're doing a lot of learnings here. So can you guys each take a minute and explain the key learnings from this, as you go through the process, certainly in the business side, this is a big imperative to do this. You want to have a business outcome that keeps your users there. But what did you learn, what was some of the learnings you guys got from the project? >> The most important learning from the company was cleaning the data, that sounds funny but, as we say in analysis, garbage in, garbage out. And that was very important for us, one of the things that we learned, that we need to put cleaning data or the system. Also, the governance. Many people forget about the governance, the governance of the data. And right now we're working, again with IBM, in order to put that governance soon. >> So data quality problem. >> Yeah, data quality. >> And, do you report into I guess, COO or the CIO, are you a peer of the CIO, how does that work? >> Oh, okay, that's another funny history because, because the company is, right now I'm working for planning. Yes, it's strange, we're working for planning for the company-- >> For business planning. >> Yeah, for business planning. >> I was coming for an engineer, engineering, and right now I'm working for planning, and trying to make money for the company. You know, that is an engineer thinking how to get more money for the company. I was talking about some kind of analytics that is geospatial analytics, and I went to see that engineer to know how their network's handling, how the quality of the network and right now introducing the same software, the same knowledge, to know which is the better points to do sales. It's a good combination where finally I'm working for planning, and my boss, the planning chief, is working for the CEO. And I hear about different organizations, somebody's in financial, the CDO's in financial, or the CDO for IT, it's different, it depends on the company. Right now, I'm working for planning, how to handle the things, how to make more money for the company, how to handle the churn, and it's interesting because all the knowledge that I have from engineering is perfect to do it. >> Well, I would argue that's the job of a CDO, is to figure out how to make money with data, or save money, right? >> Yeah. >> Yeah, absolutely. >> So it's number one, anyway, is start there. >> Yeah, the thing we always talk about it is, is really proving value, it starts with that use case, identify where the real value is, and then we can, you know, the technology can come and the development can work after that. So I agree 100% with that, is what we're seeing across the board. >> Carlos, thanks for coming in, largest telecommunications in Columbia, great customer reference. >> Carlo, take a minute to explain, real quick, get a plug in for your Data Science Elite Team. What do you guys do, how do you engage, what are some of the projects you work on? >> Right, yeah, so we're a team of about 100 data scientists worldwide, we work side by side with clients, and our job is to really understand the problem from end to end and help in all areas, from skills, tools, and technique. And we roll and prototype, in a three Agile sprints, we use an Agile methodology, about six to eight weeks, and we kind of develop a real, we call it a proof of value. It's not a MVP just yet, or POC, but at the end of the day we prove out that we can get a model, we can do some prediction, we get a certain accuracy, and it's going to add value to the organization. >> It's not a freebie, right? >> It actually is-- >> Sorry, I'm sorry. It's not a four page service, it's a freebie, right? >> Yeah, it's no cost. >> But you got to-- >> We don't like to use free, that's what-- >> But, you got to be saying-- >> It's a good lead. >> Good to discuss that-- >> Well, we don't charge, but >> Largely. >> But it, but it, it's something that clients can take advantage of, if they've got an interesting problem, they're potentially going to do some business with you guys. >> Absolutely. >> If you're the largest telecommunication provider in the country, you get a freebie, and then, the key is, you guys dig in. >> We dig in, it's practitioners, real practitioners, with the right skills, >> Yeah. >> Working on problems. >> Great sales model. >> By the way, Claro Columbia's team, they were amazing in Columbia, we had a really good time, six to eight weeks, you know, working on a problem, and those guys all loved it too, they were-- >> Thank you. >> Before they knew it, they were coding in Python and R, and they had already knew a lot of this stuff, but they're digging in with the team, and it came well together. >> This is the secret to modernization of digital transformation-- >> Yeah. >> Is having the sales process is getting, co-creating together-- >> Absolutely. >> You guys do a great job, and I think this is a trend we'll see more of, of course, TheCUBE is bringing you live coverage here in San Francisco, at Mascone North, that's where our set is. They're shutting down the streets for IBM Think 2019, here in San Francisco. More CUBE coverage after this short break, be right back. (energetic music)

>> Live from San Francisco. It's the cube covering IBM thing twenty nineteen brought to you by IBM. >> Welcome back to the live coverage here in Mosconi North in San Francisco for IBM. Think this. The cubes coverage. I'm Jeffrey David. Launching a too great guest here. Carlos. Gavel, gavel. A chief date. Officer Clara, Columbia and Carlos. See? Good. Engage your manager. IBM data Science elite team a customer of IBM country around data science. Welcome to the Cube. Thanks for joining us. Thanks for having us. So we'll hear the street, the street to shut down a i N E. Where's the big theme? Multi cloud. But it's all about the data everywhere. People trying to put end to end solutions together to solve real business problems. Date is at the heart of all this moving date around from cloud to cloud using. Aye, aye. And technology get insights out of that. So take a minute to explain your situation, but you got to try to do. >> Okay. Okay, Perfect. Right now, we're working out a lot about the business thing because we need to use the machine learning models or all the artificial intelligence toe. Take best decisions for the company. Way. We're working with Carlo in a charming mother in order to know how how come with a boy the customers left the company Because for us it's very important to maintain our our customer toe. Now, how they're how are the cables is from them. There are two facility intelligences is next selling way to do it that way. Have a lot of challenge about that because, you know, we have a lot of data, different systems, that they're running the data way need to put all the information together to run them to run the mother's. The team that Carlo is leaving right now is helping to us a lot because we WeII know how to handle that. We know howto clean the data when you have to do the right governess for the data on the IBM iniquity is very compromised with us in there in order to do that safely. That is one of the union that is very close to us right now. She was working a lot with my team in order to run the models. You saying she was doing a lot of four. I mean, over fight on right now we are trained to do it in over the system, running this park on DH that is they? They Good way that we are. We are thinking that is going to get the gold for us way Need to maintain our customers. >> So years the largest telecommunications piece Claro in Mexico for boys and home services. Is that segments you guys are targeting? Yeah, Yeah. Scope. Size of how big is that? >> Clarisa? Largest company in Colombia For telecommunication. We have maybe fifty million customers in Colombia. More than fifty percent of the market marketer also way have many maybe two point five millions off forms in Colombia. That is more than fifty percent of the customers for from services on. Do you know that it's a big challenge for us because the competitors are all the time. Tryinto take our customers on DH the charm or they'll have toe. How's the boy that and how to I hope to do their artificial intelligence to do it much learning. It's a very good way to do that. >> So classic problem and telecommunications is Charon, right? So it's a date. A problem? Yeah, but So how did it all come about? So these guys came to you? >> Yeah. They help The game does. We got together. We talked about the problem and in turn was at the top right. These guys have a ton of data, so what we did is the team got together. We have really the way to data sensibly team works is we really helped clients in three areas. It's all about the right skills, the right people, the right tools and then the right process. So we put together a team. We put together some agile approaches on what we're going to do on DH. Then we'd get started by spinning up in environment. We took some data and we took there. And there's a lot of data is terabytes of data. We took their user data way, took their use users usage data, which is like how many text, cellphone and then bill on day that we pulled all that together and environment. Then the data scientists alongside what Carlos is team really worked on the problem, and they addressed it with, you know, machine learning, obviously target. In turn, they tried a variety of models, But actually, boost ended up being one of the better approaches on DH. They came up with a pretty good accuracy about nineties ninety two. Percent precision on the model. Predicting unpredictable turn. Yeah. >> So what did you do with that? That >> that that is a very good question because the company is preparing to handle that. I have a funny history. I said today to the business people. Okay, these customers are going to leave the company. Andi, I forget about that on DH. Two months later, I was asking Okay, what happened? They say, Okay, your model is very good. All the customers goes, >> Oh, my God, What >> this company with that they weren't working with a with information. That is the reason that we're thinking that the good ways to fame for on the right toe the left because twist them which is therefore, pulls the purposes toe Montana where our customers And in that case, we lose fifty thousand customers because we didn't do nothing Where we are close in the circle, we are taking care about that prescriptive boys could have tto do it on. OK, maybe that is her name. Voice problem. We need to correct them to fix the problem in orderto avoid that. But the fetus first parties toe predict toe. Get any score for the charm on Tau handled that with people obviously working. Also at the root cause analysis because way need to charm, way, need to fix from their road, >> Carla. So walk us through the scope of, like, just the project, because this is a concern we see in the industry a lot of data. How do I attack it? What's the scoop? You just come in and just into a data lake. How do you get to the value? These insights quickly because, honestly, they're starving for insights would take us through that quick process. >> Well, you know, every every problems with different. We helped hundreds of clients in different ways. But this pig a problem. It was a big data problem because we knew we had a lot of data. They had a new environment, but some of the data wasn't there. So what we did was way spun up a separate environment. We pulled some of the big data in there. We also pulled some of the other data together on DH. We started to do analysis on that kind of separately in the cloud, which is a little different, but we're working now to push that down into their Duke Data Lake, because not all the data is there, but some of the data is there, and we want to use some of that >> computer that almost to audit. Almost figure out what you want, what you want to pull in first, absolutely tie into the business on the business side. What would you guys like waiting for the answers? Or was that some of the on your side of process? How did it go down? >> I'm thinking about our business way. We're talking a little bit about about that about their detective tow hundred that I see before data within. That is a very good solution for that because we need infested toe, have us in orderto get the answers because finally we have a question we have question quite by. The customers are leaving us. Andi. What is data on the data handed in the good in a good way with governor? Dance with data cleaning with the rhyme orders toe. Do that on DH Right now, our concern is Business Section a business offer Because because the solution for the companies that way always, the new problems are coming from the data >> started ten years ago, you probably didn't have a new cluster to solve this problem. Data was maybe maybe isn't a data warehouse that maybe it wasn't And you probably weren't chief data officer back then. You know that roll kind of didn't exist, so a lot has changed in the last ten years. My question is, do you first of all be adjusting your comment on that? But do you see a point in which you could now take remedial action or maybe even automate some of that remedial action using machine intelligence and that data cloud or however else you do it to actually take action on behalf of the brand before humans who are without even human involvement foresee a day? >> Yeah. So just a comment on your thought about the times I've been doing technology for twenty something years, and data science is something has been around, but it's kind of evolved in software development. My thought is, uh, you know, we have these rolls of data scientists, but a lot of the feature engineering Data prep does require traditional people that were devious. And now Dave engineers and variety of skills come together, and that's what we try to do in every project. Just add that comment. A ce faras predicted ahead of time. Like, I think you're trying to say what data? Help me understand >> you. You know, you've got a ninety three percent accuracy. Okay, So I presume you take that, You give it to the business businesses, Okay? Let's maybe, you know, reach out to them, maybe do a little incentive or you know what kind of action in the machines take action on behalf of your brand? Do you foresee a day >> so that my thought is for Clara, Columbia and Carlos? But but obviously this is to me. Remain is the predictive models we build will obviously be deployed. And then it would interact with their digital mobile applications. So in real time, it'll react for the customers. And then obviously, you know, you want to make sure that claro and company trust that and it's making accurate predictions. And that's where a lot more, you know, we have to do some model validation and evaluation of that so they can begin to trust those predictions. I think is where >> I want to get your thoughts on this because you're doing a lot of learnings here. So can you guys each taking minutes playing the key Learnings from this As you go through the process? Certainly in the business side, there's a big imperative to do this. You want to have a business outcome that keeps the users there. But what did you learn? What was some of the learnings? You guys gone from the project? >> They the most important learning front from the company that wass teen in the data that that sound funny, but waiting in an alley, garbage in garbage, out on DH that wass very, very important for other was one of the things that we learn that we need to put cleaning date over the system. Also, the government's many people forget about the governments of the governments of the data on DH. Right now, we're working again with IBM in our government's >> so data quality problem? Yeah, they fight it and you report in to your CEO or the CEO. Seo, your spear of the CIA is OK. That >> is it. That's on another funny history, because because the company the company is right now, I am working for planning. This is saying they were working for planning for the company. >> Business planning? >> Yeah, for business planning. I was coming for an engineer engineering on DH. Right now, I'm working for a planning on trying to make money for the company, and you know that it's an engineer thinking how to get more money for the company I was talking about. So on some kind of analysis ticks, that is us Partial Analytics on I want you seeing that in engineer to know how the network handling how the quality of the network on right now using the same software this acknowledge, to know which is the better point to do sales is is a good combination finally and working. Ralph of planning on my boss, the planning the planet is working for the CEO and I heard about different organizations. Somebody's in Financial City owes in financial or the video for it is different. That depends from the company. Right now, I'm working for planning how to handle things, to make more money for the company, how to tow hundred children. And it is interesting because all the knowledge that I have engineering is perfect to do it >> Well, I would argue that's the job of a CDO is to figure out how to make money with data. Are saying money. Yeah. Absolute number one. Anyway, start there. >> Yeah, The thing we always talked about is really proving value. It starts with that use case. Identify where the real value is and then waken. You know, technology could come in the in the development work after that. So I agree with hundred percent. >> Carlos. Thanks for coming in. Largest telecommunication in Colombia. Great. Great customer reference. Carlo thinking men to explain real quick in a plug in for your data science elite team. What do you guys do? How do you engage? What? Some of the projects you work on Grey >> out. So we were a team of about one hundred data scientists worldwide. We work side by side with clients. In our job is to really understand the problem from end and help in all areas from skills, tools and technique. And we won't prototype in a three agile sprints. We use an agile methodology about six to eight weeks and we tied. It developed a really We call it a proof of value. It's it's not a M v P just yet or or poc But at the end of the day we prove out that we could get a model. We can do some prediction. We get a certain accuracy and it's gonna add value to the >> guys. Just >> It's not a freebie. It actually sorry. I'm sorry. It's not for paint service. It's a freebie is no cough you've got. But I don't like to use >> free way. Don't charge, but >> But it's something that clients could take advantage of if they're interesting problem and maybe eventually going to do some business. >> If you the largest telecommunication provider in the country, to get a freebie and then three keys, You guys dig in because its practitioners, real practitioners with the right skills, working on problems that way. Claro, >> Colombia's team. They were amazing. In Colombia. We had a really good time. Six to eight weeks working on it. You know, a problem on those guys. All loved it, too. They were. They were. Before they knew it. They were coding and python. And are they ready? Knew a lot of this stuff, but they're digging in with the team and became well together. >> This is the secret to modernization of digital transformation, Having sales process is getting co creating together. Absolutely. Guys do a great job, and I think this is a trend will see more of. Of course, the cubes bring you live coverage here in San Francisco at Mosconi. Nor That's where I said it is. They're shutting down the streets for IBM. Think twenty here in San Francisco, more cube coverage after the short break right back.