>> Announcer: From around the globe, it's theCUBE! With digital coverage of IBM Think 2021. Brought to you by IBM. >> Well welcome everyone as theCUBE continues our IBM Think series. It's a pleasure to have you with us here on theCUBE. I'm John Walls, and we're joined today by Brian Loveys who is the Director of Offering Management for Customer and Employee Care Applications at IBM in the Data and AI Division. So, Brian, thanks for joining us from Ottawa, Canada. Good to see you today. >> Yeah, great to be here, John. And looking forward to the session today. >> Which, by the way, I've learned Ottawa are the home of the world's largest ice skating rink. I doubt we get into that today, but it is interesting food for thought. So, Brian, first off, let's just talk about the AI landscape right now. I know IBM obviously very heavily invested in that. Just in terms of how you see this currently in terms of enterprise adoption, what people are doing with it, and just how you would talk about the state of the industry right now. >> You know, it's a really interesting one, right? I think if you look at it, you know, different companies, different industries, frankly, are at different stages of their AI journey, right? I think for me personally, what was really interesting was, and we're all going through the pandemic right now, but last year with COVID-19 in the March timeframe, it was really interesting to see the impact, frankly, in the space that I play predominantly in around customer care, right? When the pandemic hit, immediately call centers, contact centers got flooded with calls, right? And so it created a lot of problems for organizations. But what was interesting to me is it accelerated a lot of adoption of AI to organizations that typically lag in technology, right? So if you think about public sector, right, that was one area that got hit very, very hard with questions and those types of things, and trying to, you know, communicate out information. So it was really interesting to see those organizations, frankly, accelerate really, really quickly, right? And if you actually, you know, talk to those organizations now, I think one of the most interesting things to me in thinking about it and talking to them now is like, hey, you know, we can do this, right? AI is really not that complicated. It can be simplified, we can take advantage of it and all of those types of things, right? So I think for me, you know, I kind of see different industries at sort of different levels, but I think with COVID in particularly, you know, and frankly not just COVID, but even digital transformation alongside COVID is really driving a lot of AI in an accelerated manner. The other thing that I'll kind of talk to a little bit here is I still think we're very much in the early innings of this, right? There's a tremendous opportunity to innovate in this space. And I think we all know that, you know, data is continually being created every single day. And as more people become even more digitalized, there's more and more data being created. Like it's how do you start to harness that data more effectively, right, in your business every day. And frankly, I think we're just scratching the surface on it. And I think tremendous amount of opportunity as we move forward. >> Yeah, you really raised an interesting point which I hadn't thought about in terms of, we think about disruptors, we think about technology being a disruptor, right, but in this case it was purely, or really largely environment, you know, that was driving this disruption, right, forcing people to make these adoption moves and transitions maybe a little quicker than they expected. Well, so because of that, because maybe somebody had to speed up their timetable for deployments and what have you, what kind of challenges have they run into then, where, because as you describe it, it's not been the more organic kind of decision-making that might be made sometimes, situation dictated it. So what have you seen in terms of challenges, you know, barriers, or just a little more complexity, perhaps, for some people who're just now getting into the space because of the environment you were talking about? >> I think a lot of this is like, you know, people don't know where to get started, right, a lot of the time, or how AI can be applied. So a lot of this is going to be about education in terms of what it can and cannot do. And then it all depends on the use cases you're talking about, right? So if I think about, you know, building out machine learning models and those types of things, right, you know, the set of challenges that people will typically face in these types of things are, you know, how do I, you know, collect all the data that I need to go build these models, right? How do I organize that data? You know, how do I get the skillsets needed to ultimately, you know, take advantage of all of that data to actually then apply to where I need it in my business, right? So a lot of this is, you know, people need to understand those concepts or those pieces to ultimately be successful with AI. And you know, what IBM is doing right here, and I'll kind of, this will be a key theme throughout this conversation today is, you know, how do you sort of lower the time to value to get there across that spectrum, but also, you know, frankly, the skills required along the way as well? But a lot of it is like, people don't know what they don't know at the end of the day. >> Well, let me ask you about your AI play then. A lot of people involved in this space, as you well know, competition's pretty fierce and pretty widespread. There's a deep bench here. In terms of IBM though, what do you see as kind of your market differentiator then? You know, what do you think sets you apart in terms of what you're offering in terms of AI deployments and solutions? >> No, that's a great question. I think it's a multifaceted answer, frankly. The first thing I'll kind of talk through a little bit, right, is really around our platform and our framework, right? We kind of refer to as our AI ladder, but it's really an integrated, you know, sort of cohesive platform for companies around the journey to AI, right? So kind of what I was mentioning a bit earlier, right? If you think about, you know, AI is really about supplying the right data into AI, and then being able to infuse it to where you need it to go, right? So to do that, you need a lot of the underlying information architecture to do that, right? So you need the ability to collect the data. You need the ability to organize the data. You need the ability to build out these models or analyze the data, right? And then of course you need to be able to infuse that AI wherever you need it to be, right? And so we have a really nice integrated platform that frankly can be deployed on any cloud, right, so we get the flexibility of that deployment model with that integrated platform. And if you think about it, we also have built, right, you know, sort of these industry-leading AI applications that sit on top of that platform and that underlying infrastructure, right? So Watson Assistant, right, our conversational AI which we'll talk probably a little bit more on this conversation, right? Watson Discovery focused on, you know, intelligent document processing, right, AI search type applications. We've got these sort of market-leading applications that sit on top, but there's also other things, right? Like we have a very, very strong research arm, right, that continues to invest and funnel innovations into our product platform and into our product portfolio, right? I think many people are aware of Project Debater we took on some of the top debaters in the world, right? But research ultimately is very much tied, right, and even, you know, some of the teams that I work with on the ground, we've got them tied directly into the squads that build these products, right? So we have this really big strong research arm that continues to bring innovation around AI and around other aspects into that product portfolio. But it's not just- >> I'm sorry go ahead, please. >> Go ahead, sorry. >> No, no, you go, (laughs) I interrupted, you go ahead. >> Don't worry, I was just going to say, the other two things I'll say like, you know, I'm saying this right, but we've got a lot of sort of proof points in around it, right, so if you talk about the scale, right, the number of customers, the number of case studies, the number of references across the board, right, in around AI at IBM it is significant, right? And not only that, but we've got a lot of, sort of I'll say industry and third-party industry recognition, right? So think about most people are aware of sort of Gartner Magic Quadrants, right, and we're the leader almost across the board, right, or a leader across the board. So, you know, cloud AI developer service, insight engines, machine learning, go down the line. So, you know, if you don't trust me, there's certainly a lot of third party validation around that as well, if that makes sense. >> Yeah, sure does. You know, we hear a lot about conversational AI and, you know, with online chat bots and voice assistance, and a myriad applications in that respect. Let's talk about conversational right now. Some people think is a little narrow, but yet there appears to be a pretty broad opportunity at the same time. So let's talk about that conversational AI element to what you're talking about at IBM and how that is coming into play. And perhaps is a pretty big growth sector in this space. >> Yeah, I think, again, I talk about scratching the surface, early innings, you'll see that theme a lot too. And I think this is another area around that, right? So, listen, let's talk about the broader side. Let's first talk about where conversational AI is typically applied, right? So you see it in customer service. That's the obvious place where I've seen the most deployments in. But if you think about, it's not just really around customer service, right? There's use cases around sales and marketing. You can think about, you know, lead qualification for example, right. You know, I'm on a website, how can I get information about a product or service? How can I automate some of that information collection, answering questions, how can I schedule console? All those things can be automated using, right, conversational AI, but organizations don't want these sort of points solutions across the customer journey. What they're ultimately looking for is a single assistant to kind of, you know, front that particular customer. So what if I do come on from a lead qual perspective, but really I'm not there for lead qual, I'm actually a customer, and I want to get a question answered, right? You don't want to have these awkward starts and stops with organizations, right? So on the customer side where we see the conversational AI going is really sort of covering that whole gambit in terms of that customer journey, right? And it's not just the customer journey, but you also want to be across channels, right? So you can imagine not just, you know, the website and the chat on the website, but also, right, across your messaging channels, across your phone, right? And not just that, but you also want to be able to have a really nice experience around, hey maybe I'm on a phone call with some automation, but I need to be able to hand them off to a digital play, right? Maybe that's easier to sign up for a particular offer, or do some authentication, or whatever it might be, right? So to sort of be able to switch between the channels is really, really going to become more important in terms of a seamless experience as you do kind of go through it, right- >> So let's talk about customers- >> Oh, go ahead sir. >> Yeah, you talked about customers a little bit, and you mentioned case studies, but I hope we can get into some specifics, if you can give us some examples about people, companies with whom you've worked and some success that you've had in that respect. And I think maybe the usual suspects come to mind. I think about finance, I think about healthcare, but you said, "Hey buddy, but customer call issues, you know, service centers, that kind of thing would certainly come into play," but can you give us an idea or some examples of deployments and how this is actually working today? >> Oh, absolutely, right? So I think you were kind of mentioning, you were talking about sort of industries that are relevant, right? So, you know, the ones that I think are most relevant that we've seen are the ones with the biggest sort of consumer side of it, right? So clearly in financial services, banks, insurance are clearly obvious ones. Telecommunication, retail, healthcare, these are all sort of big industries with a lot of sort of customers coming in, right? And so you'll see different use cases in those industries as well, right? So the obvious one, we've got a really good client, Royal Bank of Scotland, they've now changed their name to NatWest in Scotland. So they started out with customer service, right? So dealing with personal banking questions through their website. What's interesting, and you'll see this with a lot of these use cases is they will start small, right, with a single use case, but they'll start to expand from there. So for example, NatWest, right, they're starting with personal banking, but they're now expanding to other areas of the business across that customer journey, right? So that's a great example of where we've seen it. Cardinal Health, right, because we're not dealing with customers in terms of external customers, but dealing with internal customers, right, from an IT help desk standpoint. So it's not always external customers. Oftentimes, frankly, it can be employees, right? So they are using it through an IDR system, right? So through over the phone, right, so I can call, instead of getting that 1-800 number, I'm going to get a nice natural language experience over the phone to help employees with common problems that they have with their help desk. So, and they started really, really small, right? They started with, you know, simple things like password resets, but that represented a tremendous amount of volume that ultimately hit at their call centers. So NatWest is a great example. CIBC, another bank in Canada, Toronto, is a great example. And the nice thing about what CIBC is doing and they're a big, you know, we have four big banks here in Canada. What CIBC do is really focusing a lot on the transactional side. So making it really easy to do interact transfers or send money, or all those types of things, or check your balance or whatever it might be. So putting a nice, simple interface on some of those common, transactional things that you would do with a bank as well. >> You know, before I let you go, I'd like to hit just a buzzword we hear a lot of these days, natural language processing, NLP. All right, so NLP, define that in terms of how you see it and how is it being applied today? Why does NLP matter, and what kind of differences is it making? >> Wow, natural language processing is a loaded term as a buzzword, I completely agree. I mean, listen, at the 50,000 foot level, natural language processing is really about understanding language, right? So what do I mean by that? So let's use the simple conversational example we just talked about. If somebody's asking about, you know, "I'd like to reset my password," right? You have to be able to understand, well what is the intent behind what that user is trying to do, right? They're trying to reset a password, right? So being able to understand that inquiry that user has that's coming in and being able to understand what the intent is behind it. That's sort of one key aspect of natural language processing, right? What is the intent or the topic around that paragraph or whatever it might be. The other sort of key thing around natural language processing, the importance of extracting certain things that you need to know. And again, using the conversational AI side, just for a minute, to give a simple example. If I said, "You know what, I need to reset my password." I know what the intent is, I want to reset a password, but, right, I don't know which password I'm trying to reset. Right, and so this is where sort of you have to be able to extract objects, and we call them entities a lot of the time and sort of the (indistinct) or lingo. But you got to be able to extract those elements. So, you know, I want to reset my ATM password. Great, right, so I know what they're trying to do, but I also need to extract that it's the ATM password that I'm trying to do. So that's one sort of key angle, natural language processing, and there's a lot of different AI techniques to be able to do those types of things. I'll also tell you though, there's a lot around the content side of the fence as well. So you can imagine how like a contract, right, and there were thousands of these contracts, and some of your terms may change. You know, how do you know, out of those thousands of contracts where the problems are, where I need to start looking, right? So another sort of key area of natural language processing is looking at the content itself, right? Can I look at these contracts and automatically understand that this is an indemnity clause, right? Or this is an obligation, right? Or those types of things, right, and being able to sort of pick those things out, so that I can help deal with those sort of contract-processing things. So that's sort of a second dimension. The third dimension I'll kind of give around this is really around, you can think about extracting things like sentiment, right? So we talked about, you know, extracting objects and nouns, and those types of things, but maybe I want to know in an analytics use case with customers, you know, what is the sentiment and, you know, analyzing social media posts or whatever it might be, what's the sentiment that people have around my product or service. So natural language process, if you think about it at the real high level is really about how do I understand language, but there's a variety of sort of ways to do that, if that makes sense. >> Yeah, no sure, and I think there are a lot of people out there saying, "Yeah, the sooner we can identify exasperation (laughs) the better off we're going to be, right, in handling the problems." So, it's hard work, but it's to make our lives easier, and congratulations for your fine work in that space. And thanks for joining us here on theCUBE. We appreciate the time today, Brian. >> Thank you very much. >> You bet, Brian Loveys, he's talking to us from IBM, talking about conversational AI and what it can do for you. I'm John Walls, thanks for joining us here on theCUBE. (upbeat music) ♪ Dah, deeah ♪ ♪ Dah, dee ♪ (chimes ringing)

>>from >>Around the globe. It's the cube with digital coverage of IBM think 2021 brought to you by IBM >>Well welcome everyone is the cube continues or IBM Thanks series. It's a pleasure to have you with us here on the cube. I'm john walls and we're joined today by brian loves who is the director of offering management for customer and employee care applications in the at IBM in the data and AI division. So brian, thanks for joining us from Ottawa Canada, good to see you today. >>Yeah, great to be here john I'm looking forward to the session today >>which by the way I've learned Ottawa is the home of the world's largest ice skating rink. I doubt we'll get into that today, but it is interesting food for thought. Uh so brian first off, let's just talk about um the Ai landscape right now. I know IBM obviously very heavily invested in that uh just in terms of how you see this currently as in terms of enterprise adoption, what people are doing with it and and just how you would talk about the state of the industry right now, >>you know, it's a really interesting one, right? I think if you look at it, you know different companies, different industries frankly are at different stages of their Ai journey, right? Um I think for me personally what was really interesting was, and we're all going through the pandemic right now, but last year with covid 19 in the March timeframe, it was really interesting to see the impact, frankly in the space that I played predominantly in around customer care, right? When the pandemic hit immediately call centers, contact centres got flooded with calls, right? And so it created a lot of problems for organizations. But it was interesting to me is it accelerated a lot of adoption of ai to organizations that typically lag and technology. Right? So if you think about public sector, right, that was one area that got hit very, very hard with questions and those types of things and trying to communicate and communicate out information. So it was really interesting to see those organizations frankly accelerate really, really quickly, right? And if you actually talk to those organizations now, I think one of the most interesting things to me and thinking about it and talking to them now is like, hey, you know, we can do this right, AI is really not that complicated, it can be simplified, we can take advantage of it and all of those types of things. Right? So I think for me, you know, I kind of see different industries that sort of different levels, but I think with Covid in particularly, you know, and frankly not just Covid, but even digital transformation alongside Covid is really driving a lot of ai in an accelerated manner. The other thing I'll kind of I'll kind of talk to a little bit here is I still think we're very much in the early innings of this, right, there is a tremendous opportunity innovating in the space and I think we all know that you know data is continually being created every single day and as more people become even more digitalized, there's more and more data being created. Like how do you start to harness that data more effectively, right in your business every day? And frankly I think we're just scratching scratching the surface on it and I think tremendous amount of opportunity as we move forward. >>Yeah, he really is really raised an interesting point which I hadn't thought about in terms of, we think about disruptors, we think about technology being a disrupter, right? But in this case it was purely really, largely environment that was driving this disruption, right, forcing people to to make these adoption moves and transitions maybe a little quicker than they expected. So because of that, because maybe somebody had to speed up their timetable for deployments and what have you what what kind of challenges have they run into them? Where because, as you describe it, it's not been the more organic kind of decision making that might be made, sometimes situation dictated it. So what have you seen in terms of challenges, barriers or just a little more complexity perhaps for some people who are just not getting into the space because of the environment you were talking about? >>I think a lot of this is like people don't know where to get started, right, a lot of the time or how ai can be applied. So a lot of this is going to be a bad education in terms of what it can and cannot do, and then it all depends on the use cases you're talking about, right? So if I think about, you know, building a machine learning models and those types of things right? You know, this set of challenges that people will typically face in these types of things are, you know, how do I collect all the data that I need to go build these models? Right? How do I organize that data? Um you know, how do I get the skill sets needed to ultimately, you know, take advantage of all that data to actually then apply to where I needed in my business? Right, So a lot of this is, you know, people need to understand, you know, those concepts are those pieces um to ultimately be successful with AI and you know what IBM is doing right here and I'll kind of this will be a key theme through this conversation today, is how do you sort of lower the time to value, to get there across that spectrum, but also, you know, frankly the skills >>required along the way as >>well, but a lot of it is like people don't know what they don't know at the end of the day. Mhm. >>Well, let me ask you about about your AI play then, a lot of people involved in this space, as you well know, you know, competitions pretty fierce and pretty widespread, there's a deep bench here um in terms of IBM know, what do you see is kind of your market different differentiator then, you know, what what do you think set you apart in terms of what you're offering in terms of AI deployments and solutions? >>No, that's a great question. I think it's a multifaceted answer, frankly. Um the first thing I'll kind of talk through a little bit right, is really around our platform and our our framework, right? We could refer to as our air ladder, um but it's really an integrated, you know, sort of cohesive platform for companies around the journey to AI, right? So kind of what I was mentioning earlier, right? If you think about, you know, AI is really about supplying the right data into A I. And then being able to infuse it to where you needed to go. Right? So to do that, you need a lot of the underlying information architecture to do that, Right? So you need the ability to collect the data, you need the ability to organize the data, you need the ability to to build out these models, right? Or analyze the data and then of course you need to be able to infuse that ai wherever you need it to be. Right. And so we have a really nice integrated platform that frankly can be deployed on any cloud. Right? So we got the flexibility that deployment model with that in greater platform. And you think about it? We also have built right, you know, sort of these industry leading Ai applications that sit on top of that platform and that underlying infrastructure. Right? So Watson assistant, Right. Our conversational AI, which we'll talk probably a little bit more on this conversation. Right, Watson discovery focus on, you know, intelligent document processing, right. AI search type applications. We've got these sort of market leading applications that sit on top, but there's also other things, right? Like we have a very, very strong research arm right, that continues to invest and funnel innovations into our product platform and into our product portfolio. Right? I think many people are aware of project debater, we took on some of the top debaters in the world, right? But research ultimately is very much tied, right? And even some of the teams that I work with on the ground, we've got them tied directly into the squads that build these products, Right? So we have this really big strong research arm that continues to bring innovation around AI and around other aspects into that product portfolio. But it's not just go ahead, >>Please go ahead. three. No, no. You know, I interrupted you. Go ahead. >>No, I was just gonna say that the other two things, I'll say it like, you know, I'm saying this right, but we've got a lot of sort of proof points and around it. Right? So, if you talk about the scale right? The number of customers, the number of case studies, a number of references across the board, right? In around AI AT IBM It is significant, Right? Um, and not only that, but we've got a lot of sort of, I'll say industry and third party industry recognition. Right? So think about most people are aware of sort of Gartner magic quadrants, right? And we're the leader almost across the board, Right? Or a leader across the board. So cloudy I developer service inside engines, machine learning go down the line. So, you know, if you don't trust me, there's certainly a lot of third party validation around that as well. That makes sense. >>Yeah, it sure does. You know, we're hearing a lot about conversational AI and, you know, with online chat bots and voice assistance and a myriad applications in that respect. Let's talk about conversational right now. Some people think it's little narrow, but, but yet there appears to be a pretty broad opportunity at the same time. So let's talk about that conversational AI um, uh, element um, to what you're talking about at IBM and how that is coming into play and, and perhaps is a pretty big growth sector in this space. >>Yeah, I think again, I talked about scratching the surface early innings. You'll see that theme a lot too. And I think this is another area around that. So listen, let's talk about the broader side. Let's first talk about where conversation always typically applied. Right? So you see it in customer service, that's the obvious place we're seeing the most appointments in. But if you think about, it's not just really around customer service, right? There's use cases around sales and marketing. If you think about, you know, lead qualification, for example, right? How can, you know, I'm on a website, how can I get information about a product or service? How can I automate some of that information collection, answering questions? How can I schedule console? All those things can be automated using great conversationally. I, the organizations don't want these sort of point solutions across the customer journey. What we're ultimately looking for is a single assistant to kind of, you know, front right, that particular customer. So what if I do come on from a legal perspective, but really I'm not here for legal. I'm actually a customer and I want to get a question answered, right? You don't want to have these awkward starts and stops with organizations, Right? So on the customer side where we see the conversation like, hey, I going and it's really kind of covering that full gambit in terms of that customer journey, right? And it's not just the customer journey, but you also want to be across channels, right? So you can imagine right now, not just, you know, the website and the chat on the website, but also right across their messaging channels, right across your phone. Right. And not just that, but you also want to be a really nice experience around, hey, maybe I'm on a phone call with some automation, but I need to be able to hand them off to a digital play. Right? Maybe that's easier to sign up for a particular offer or do some authentication or whatever might be, right. So to sort of be able to sort of switch between the channels, it's really, really going to become more important in this sort of sort of seamless experience as you just kind of go through it. Right? >>So you're coming by customers. Yeah. >>You talked about customers a little bit and you mentioned case studies, but can we get, I hope we can get into some specifics. You can give us some examples about people, companies with whom you've worked and and some success that you've had that respect. And I think maybe the usual suspects come to mind about finance. I might health care, but you said anybody with customer call issues, service centers, that kind of thing would certainly come into play. But can you give us an idea or some examples of deployments and how this is actually working today? >>Oh, absolutely. Right. So I think you kind of mentioned you become sort of industries that are relevant. Right? So, you know, the ones that I think are most relevant that we've seen are the ones with the biggest sort of consumer sort of side to it. Right? So clearly in financial services, banks, insurance, and clearly obvious ones telecommunications, retail, healthcare, these are all sort of big industries with a lot of sort of customers coming in. Right? So you'll see different use cases in those industries as well. Right. So the obvious one, we've got a really good client, Royal Bank of Scotland, they've now changed their name to natwest Open Scotland. Um So they started out with customer service. Right? So dealing with personal banking questions through their website, what's interesting and you'll see this with a lot of these use cases is they will start small, right with a single use case that they'll start to expand from there. So, for example, >>natwest right there, starting with they started with personal banking, but they're not expanding to other areas of the business across that customer journey. Right. So it's a great example of where we've seen it. Cardinal Health Right. We're not dealing with customers in terms of external customers but dealing with internal customers right from the help that standpoint. So it's not always external customers. Oftentimes frankly it can be employees. Right? So they are using it right through an I. V. R. System. Right? So through over the phone. Right. So I can call instead of getting that 1 800 number. I'm going to get a nice natural language experience over the phone to help employees with common problems that they have with their health does so. And they started really, really small, right? They started with simple things like password resets but that represented a tremendous amount of volume but ultimately headed their cost cost centers. So not West is a great example. C I B C. Another bank in Canada Toronto is a great example and the nice thing about what CNBC is doing and there are big, you know, we have four big banks here in Canada, what have you seen do is really focusing a lot on the transactional side. So making it really easy to do interact transfers or send money or over those types of things or check your balance or whatever it might be. So putting a nice simple interface on some of those common transactional things that you >>would do with the bank as well, >>you know, before I let you go, uh I'd like to hit this of buzz where we hear a lot of these days natural language processing. NLP Alright, so, so NLP define that in terms of how you see it and and how is it being applied today? Why why does NLP matter? And what kind of difference is it making? >>Wow, that's a loaded natural language processing. There's a loaded term in a buzzword. I completely agree. I mean listen, at the 50,000 ft level, natural language processing is really about understanding length, Right? So what do I mean by that? So let's use the simple conversational example. We just talked about if somebody is asking about, I'd like to reset my password right? You have to be able to understand what is the intent behind what that user is trying to do right there? Trying to reset a password, right? So being able to understand that inquiry that the user has that's coming in and being able to understand what the intent is behind it. >>That's sort of one, you know, aspect of natural language processing, right? What is the intent or the topic around that paragraph or whatever it might be. The other sort of key thing around natural language processing the importance, extracting certain things that you need to know. And again using the conversational ai side, just for a minute to give a simple example if I said you know what I need to reset my password, I know what the intent is. I want to reset a password but Right I don't know which password I'm trying to reset. Right? So this is where you have to be able to extract objects and we call them entities a lot of time in sort of the ice bake or lingo but you've got to be able to extract those elements. So you know I want to reset my A. T. M. Password. Great. Right so I know what they're trying to do but I also need to extract that it's the A. T. M. Password that I'm trying to do. So that's one sort of key angle of natural language processing and there's a lot of different techniques to be able to do those types of things. I'll also tell you though there's a lot around the content side of the fence as well, right? So you can imagine having a contract, right? And there are thousands of these contracts and some of your terms may change. How do you know, out of those thousands of contracts where the problems are, where I need to start looking, Right? So another sort of keep key area of natural language processing is looking at the content itself. Can I look at these contracts and automatically understand that this is an indemnity clause, Right? And this is an obligation, right? Or those types of things, right? And be able to sort of pick pick those things out so that I can help deal with those sort of contract processing things. That's sort of a second dimension. The third dimensional kind of kind of give around this is really around. You can think about extracting things like sentiment, right? So we talked about, you know, extracting objects and downs and those types of things. But maybe I want to know and analytics use case with customers. Um you know, what is the sentiment and you know, analyzing social media posts or whatever it might be. What's the sentiment that people have around my product or service? So naturally this process, if you think about it, the real high level is really about how do I understand language? But there's a variety of sort of ways to do that if that makes sense? >>Yeah, sure. And I think there's a lot of people out there saying, yeah, the sooner we can identify exasperation, the better off we're going to be right and handling the problems. But it's hard work but it's to make our lives easier and congratulations for your fine work in that space. And thanks for joining us here on the cube. We appreciate the time. Today, brian, >>thank very much. >>You bet BRian Levine is talking to us from IBM talking about conversational Ai and what it can do for you. I'm john Walsh, thanks for joining us here on the cube. Mhm. >>Mhm.

(upbeat techno music) >> Announcer: From New York City, it's theCUBE, covering New Relic FutureStack 2019, brought to you by the New Relic. >> Hi, I'm Stu Miniman and this is theCUBE's exclusive coverage of New Relic's Futurestack 2019 here in New York City, seventh year of the show. Our first year here, about 600 or so in attendance, and real excited, because we've had some of the users here to help kick off our coverage. And joining us, first time guest on the program, Josh Biggely is a senior engineer of Enterprise Monitoring, with Cardinal Health coming to us from a little bit further north and east than I do, Prince Edward Island, thank you so much for coming here to New York City and joining me on the program. >> Yeah, thanks for having me Stu, I'm excited to be here. I haven't been in New York, it's probably been more two decades. So it's nice to be back in a big city, I live in a very small place. >> Yeah, so if you go to Times Square, it's now Disneyland, is what we call it. It's not the 42nd street that it might've been a couple of decades ago. I grew up about 45 minutes from here, so it's gone through a lot, love the city, especially gorgeous weather we're having here in the fall. >> I'm excited for it. >> All right, so Josh, Cardinal Health, health is in the name so we think we understand a little bit about it, but tell us a little bit about the organization itself and how it's going through changes these days. >> Sure, so Cardinal Health is a global healthcare solutions provider. We are essential to care, which means we deliver the products and solutions that your healthcare providers need to literally cure disease, keep people healthy. So we're in 85% of the hospitals in the United States, 26,000 pharmacies, about 3,000,000 different home healthcare users receive products from us. Again we're global, so we're based in Dublin, Ohio, just outside of Columbus. But obviously, I live in Canada so I work for the Cardinal Health Canada Division. We've got acquisitions around the world. So yeah, it's an exciting company. We've recently gone through a transformation not only as a company, but from a technology side where we've shifted one of our data centers entirely into the cloud. >> All right, and Josh, your role inside the company, tell us a little bit about, you said it's global, what's under your purview? >> So my team is responsible for Enterprise Monitoring, and that means that we develop, deploy, support and integrate solutions for monitoring both infrastructure applications and digital experience for our customers. We have a number of tools, including New Relic, that we use. But it's a broad scope for a small team. >> Stu: Okay, and you've talked about that transformation. Walk us through a little bit about that, what led to, as you said, some big moves into public cloud? >> Yeah, our team is part of an overall effort to allow Cardinal Health to be more adaptive, to be more agile. The move to cloud allows teams that are developing applications and platforms to make a decision how to respond to the needs of their customers more rapidly. Gone are the days of, "I need a new server, "I need to predict six months from now "that I'm going to need a new server, "put the order in, get it delivered, "get it racked, get it wired." We watch a lot of people, the provision on demand. I mean, our senior vice president, or my senior vice president, likes to say, "I want you to fail fast, fail cheap." He does not say fail often. Although sometimes I do that, but that's okay. As long as you recognize that you're failing and can roll that back, redeploy, It's been really transformative for my team in particular, who was very infrastructure focused when I started with the company five years ago. >> Stu: All right, and can you bring us inside from your application portfolio, was it a set of applications, was it an entire data center? What moved over, how long did it take, and can you share what cloud you're using? >> Sure, so it's been about a two year journey. We're actually a multicloud company. We've got a small footprint in Azure, small footprint in AWS, but we're primarily in Google Cloud. We are shutting down one data center, we are minimizing another data center, and we've moved everything. We've moved everything from small bespoke applications that are targeted on team to entire ecommerce platforms and we've done everything from lift and shift, which I know you don't like to hear. But we've done lift and shift, we've done rehosting, we've done refactoring and we have re-architected entire platforms. >> Yeah, so if you could expand a little bit when we say lift and shift, I'm fine with lift and shift as long as there's another word or plan after that which I'm expecting you do have. >> Josh: Yeah, absolutely. So the lift and shift was, "Hey, let's move from our data centers into GCP. "Let's give teams the visibility, the observability "that they need so that they can make the decisions on "what they need to do best." In a lot of cases, or in fact, in 15% of the 6,500 severs that we touch, we actually full out decommed the instance. Teams had them, they were running at our data centers but they weren't actually providing any value to the company. >> So you said your team before was mostly concerned about infrastructure and a lot of what you did is now on GCP so you fired the entire team and you hired a bunch of PhDs to be able to manage Google environments? >> Absolutely not. (laughter) The principals of enterprise monitoring as a practice still apply in a cloud. We are, at heart, data geeks. And I would fair say that we're actually data story tellers. Our job is to give tools and methodologies to application teams who know what the data means in context, but we give the tools to provide that data to them. >> Stu: All right, love that. I believe I've actually seen data geek shirts at the the New Relic shows itself. But data story tellers, that was kind of thing that you heard, "I have a data scientist "that's going to help us to do this." Is that data scientist in New York or are you actually enabling who is able to tell those data stories today? >> So that is the unique part. Data story telling is not a data science. I wish that I could be a data scientist, I like math, but I'm not nearly that good at it. A data story teller takes the data and the narrative of the business, and weaves them together. When you tell someone, "Here's some data." They will look at it and they will develop their own narrative around it. But as a story teller you help craft that narrative for them. They're going to look at that data and they're going to feel it, They're going to understand it and it's going to motivate them to act in a way that is aligned with what the business objectives are. So data story tellers come in all forms. They come as monitoring engineers, they're app engineers, but they're also people who are facing the customer, they're business leaders, they're people in our distribution centers who are trying to understand the impacts of orders in their order flow, in their personnel that they have. It is a discipline that anyone can engage in if we're willing to give them the right tools. >> All right, so Josh, you got rid of a data center, you're minimizing a data center, you're shifting to cloud, you're making a lot of changes and now being able to tell data stories. Can you tell us organizationally everything goes smoothly or are their anythings that you learned along the way that maybe you could share with your peers to help them along that journey? And any rough spots, with hindsight being what it is, that you might be able to learn from? >> Yeah, so hindsight definitely 20/20. The one thing that I would say to folks is get your data right. Metadata, trusting your data is key, it's absolutely vital. We talk a lot about automation and automation is one of those things that the cloud enables very nicely. If you automate on garbage data, you are going to automate garbage generation. That was one of our struggles but I think that every organization struggles with data fidelity. But teams need to spend more time in making sure that their data, specifically their metadata, around, "Hey is this prod, is it non-prod, "what stack is this running, who built it?" Those things definitely need to be sorted out. >> Okay, talk about the observability and the monitoring that you do, how long have you been using New Relic and what products? And tell us a little about that journey. >> Sure, so we've been using New Relic for about two years. It was a bit of a slow run up to its adoption. We are a multi-tool company so we have a number of tools. Some of them are focused primarily on our network infrastructure, our on-prem storage. Although Cardinal had moved predominantly to the cloud, we have distribution centers, nuclear pharmacies all around the world. And those facilities have not gone into the cloud. So you've got network connectivity. New Relic for us has filled our cloud niche and observability, as Lou announced, is going to give us context to things that we're after. You hear the term dark data, we call them obs logs. It's data that we want to have, we only need it for a very short period of time to help us do post-op or RCAs as well as to look at, overall in our organization, the performance of the applications. For us, New Relic is going to give us an option to put data for observability. Observability is really about high fidelity data. In its world of cloud, everyone wants everything right now. And they also want it down to the millisecond. A platform that can pull that off, that's a remarkable thing. >> Yeah, Veruca Salt had it right, "I want it now." So are you using New Relic One yet? >> We have been using New Relic One for at least a couple of months going back into March this year. It's exciting, we're one of those companies that Lou talked about in his key note, we have hundreds of sub accounts. And we did so very intentfully, but it was a bit of a nightmare before we got to New Relic One. That ability for a platform team to see across multiple sub accounts, really powerful. >> Okay, so you saw a lot of announcements this morning. Anything particular that jumped out, you were excited? Because Lou kept saying over and over, and if you're using New Relic One, "This is free, this is free, this is free." That platform where it's all available for you now. >> I think the programmability is one of the things that really got me excited. One of the engineers on my team had a chance to go and sit with Lou and team, two weeks ago, and was part of that initial Hackathon. Made some really interesting things. That's exciting so shout out to Zack and the work he did. Logging, for me, is something that is huge. I know we've got data that we should have in context. So that Lou announced five terabytes of ingestion for free, all I could do was tap my fingers together and think, "Oh, okay. You're asking for it, Lou. Challenge accepted." (laughter) >> Stu: That's exciting, right. So you feel that you're going to be building apps, it sounds like already, at the FutureHack. That you're starting to move down that path. >> Definitely, and I'm really excited. Not to necessarily give it to my team. We build the patterns for teams that needs patterns, but there are so many talented individuals at Cardinal Health who, if we give them the patterns to follow, they're just going to go execute. Open sourcing that is a brilliant idea and really crowd sourcing development is the way to go. >> Yeah, I think you bring up a really interesting point. So even though your team might be the one that provides the platform, you're giving that programmability, sensibility to a broader audience inside the team and democratizing the data that you have in there. >> Yes, you keyed in on one of the things I love to talk about which is democratized access to data. Over and over again you'll hear me preach that, "I know what I know but I also know what I don't know "and more particular I don't know what I don't know. "I need other people to help me recognize that." >> We've really talked about that buzzword out there about digital transformation. When it is actually being happened, it goes from, "Oh, somebody had an opinion," to, "Wait, I actually now can actually get to the data, "and show you the data and leverage the data "to be able to take good actions on that." >> That's right, data driven decision making is not just just an idiom. It's not something that is a buzzword, it is a practice that we all need to follow. >> Stu: All right, so Josh, you're speaking here at the show. Give our audience just a quick taste, if you will, about what you're going to be sharing with your peers here at the show. >> We've actually talked about a lot of it already so I hope that people are not going to watch this session before my session later. But it really is around the power of additional transformation, the power of observability, what happens when you do things right, and the way the cloud makes teams more nimble. I won't give you it all because then people won't watch my session on Replay but, yeah, it'll be good. >> Well, definitely they should check that out. I'm hoping New Relic has that available on Replay. Give the final word here, what you're really hoping to come out of this week. Sounds like your team's deeply engaged, you've done the Hackathon, you're working with the executive teams. So FutureStack 2019, what are you hoping to walk away with? >> For me, it's about developing patterns. My team, in addition to our enterprise architecture team, is responsible for mapping out what we're going to do and how we're going to do it. Teams want to go fast and if we're not going to lay down the foundation for them to move quickly, especially in the realm of enterprise monitoring, they're going to try do it themselves. Which may or may not work. We don't want to turn teams away from using specific tools if it fits, but if there's a platform that will allow them to execute and to keep all that data centralized, that is really the key to observability. Having that high fidelity data, but then being able to ask questions, not just of the data you put in, but the data that put in maybe by a platform team or by a team that supported Kubernetes or PCF. >> All right, well, Josh Biggely, thank you so much for sharing all that you've been going through in Cardinal Health's transformation. Great to talk to you. >> Thanks so much, Stu. >> All right, lots more here at New Relic's FutureStack 2019. I'm Stu Miniman and as always, thank you for watching theCUBE. (light techno music)