>>From around the globe. It's the cube with digital coverage of an AI ops virtual forum brought to you by Broadcom. >>Welcome to the AI ops virtual forum. Finally, some Artan extended to be talking with rich lane now, senior analyst, serving infrastructure and operations professionals at Forrester. Rich. It's great to have you today. >>Thank you for having me. I think it's going to be a really fun conversation to have today. >>It is. We're going to be setting the stage for, with Richard, for the it operations challenges and the need for AI ops. That's kind of our objective here in the next 15 minutes. So rich talk to us about some of the problems that enterprise it operations are facing now in this year, that is 2020 that are going to be continuing into the next year. >>Yeah, I mean, I think we've been on this path for a while, but certainly the last eight months has, uh, has accelerated, uh, this problem and, and brought a lot of things to light that, that people were, you know, they were going through the day to day firefighting as their goal way of life. Uh, it's just not sustainable anymore. You a highly distributed environment or in the need for digital services. And, you know, one of them has been building for a while really is in the digital age, you know, we're providing so many, uh, uh, the, the interactions with customers online. Um, we've, we've added these layers of complexity, um, to applications, to infrastructure, you know, or we're in the, in the cloud or a hybrid or multi-cloud, or do you know you name it using cloud native technologies? We're using legacy stuff. We still have mainframe out there. >>Uh, you know, the, just the, the vast amount of things we have to keep track of now and process and look at the data and signals from, it's just, it's a really untenable for, for humans to do that in silos now, uh, in, in, you know, when you add to that, you know, when companies are so heavily invested in gone on the digital transformation path, and it's accelerated so much in the last, uh, year or so that, you know, we're getting so much of our business in revenue derived from these services that they become core to the business. They're not afterthoughts anymore. It's not just about having a website presence. It's, it's about deriving core business value from the services you're providing to your, through your customers. And a lot of cases, customers you're never going to meet or see at that. So it's even more important to be vigilant. >>And on top of the quality of that service that you're giving them. And then when you think about just the staffing issues we have, there's just not enough bodies to go around it in operations anymore. Um, you know, we're not going to be able to hire, you know, like we did 10 years ago, even. Uh, so that's where we need the systems to be able to bring those operational efficiencies to bear. When we say operational efficiencies, we don't mean, you know, uh, lessening head count because we can't do that. That'd be foolish. What we mean is getting the head count. We have back to burping on and higher level things, you know, working on, uh, technology refreshes and project work that that brings better digital services to customers and get them out of doing these sort of, uh, low, uh, complexity, high volume tasks that they're spending at least 20%, if not more on our third day, each day. So I think that the more we can bring intelligence to bear and automation to take those things out of their hands, the better off we are going forward. >>And I'm sure those workers are wanting to be able to have the time to deliver more value, more strategic value to the organization, to their role. And as you're saying, you know, was the demand for digital services is spiking. It's not going to go down and as consumers, if w if we have another option and we're not satisfied, we're going to go somewhere else. So, so it's really about not just surviving this time right now, it's about how do I become a business that's going to thrive going forward and exceeding expectations that are now just growing and growing. So let's talk about AI ops as a facilitator of collaboration, across business folks, it folks developers, operations, how can it facilitate collaboration, which is even more important these days? >>Yeah. So one of the great things about it is now, you know, years ago, have I gone years, as they say, uh, we would buy a tool to fit each situation. And, you know, someone that worked in network and others who will somebody worked in infrastructure from a, you know, Linux standpoint, have their tool, somebody who's from storage would have their tool. And what we found was we would have an incident, a very high impact incident occur. Everybody would get on the phone, 24 people all be looking at their siloed tool, they're siloed pieces of data. And then we'd still have to try to like link point a to B to C together, you know, just to institutional knowledge. And, uh, there was just ended up being a lot of gaps there because we couldn't understand that a certain thing happening over here was related to an advantage over here. >>Um, now when we bring all that data into one umbrella, one data Lake, whatever we want to call it, a lot of smart analytics to that data, uh, and normalize that data in a way we can contextualize it from, you know, point a to point B all the way through the application infrastructure stack. Now, the conversation changes now, the conversation changes to here is the problem, how are we going to fix it? And we're getting there immediately versus three, four or five hours of, uh, you know, hunting and pecking and looking at things and trying to try to extrapolate what we're seeing across disparate systems. Um, and that's really valuable. And in what that does is now we can change the conversation for measuring things. And in server up time and data center, performance metrics as to how are we performing as a business? How are we overall in, in real time, how are businesses being impacted by service disruption? >>We know how much money losing per minute hour, or what have you, uh, and what that translate lights into brand damage and things along those lines, that people are very interested in that. And, you know, what is the effect of making decisions either brief from a product change side? You know, if we're, we're, we're always changing the mobile apps and we're always changing the website, but do we understand what value that brings us or what negative impact that has? We can measure that now and also sales, marketing, um, they run a campaign here's your, you know, coupon for 12% off today only, uh, what does that drive to us with user engagement? We can measure that now in real time, we don't have to wait for those answers anymore. And I think, you know, having all those data and understanding the cause and effect of things increases, it enhances these feedback loops of we're making decisions as a business, as a whole to make, bring better value to our customers. >>You know, how does that tie into ops and dev initiatives? How does everything that we do if I make a change to the underlying architectures that help move the needle forward, does that hinder things, uh, all these things factor into it. In fact, there into the customer experience, which is what we're trying to do at the end of the day, w w whether operations people like it or not, we are all in the customer experience business now. And we have to realize that and work closer than ever with our business and dev partners to make sure we're delivering the highest level of customer experience we can. >>Uh, customer experience is absolutely critical for a number of reasons. I always kind of think it's inextricably linked with employee experience, but let's talk about long-term value because as organizations and every industry has pivoted multiple times this year and will probably continue to do so for the foreseeable future, for them to be able to get immediate value that let's, let's not just stop the bleeding, but let's allow them to get a competitive advantage and be really become resilient. What are some of the, uh, applications that AI ops can deliver with respect to long-term value for an organization? >>Yeah, and I think that it's, you know, you touched upon this a very important point that there is a set of short term goals you want to achieve, but they're really going to be looking towards 12, 18 months down the road. What is it going to have done for you? And I think this helps framing out for you what's most important because it'd be different for every enterprise. Um, and it also shows the ROI of doing this because there is some, you know, change is going to be involved with things you're gonna have to do. But when you look at the, the, the longer time horizon of what it brings to your business as a whole, uh it's to me, at least it all seems, it seems like a no brainer to not do it. Um, you know, thinking about the basic things, like, you know, faster remediation of, of, uh, client impacting incidents, or maybe, maybe even predictive of sort of detection of these incidents that will affect clients. >>So now you're getting, you know, at scale, you know, it's very hard to do when you have hundreds of thousands of optics of the management that relate to each other, but now you're having letting the machines and the intelligence layer find out where that problem is. You know, it's not the red thing, it's the yellow thing. Go look at that. Um, it's reducing the amount of finger pointing and what have you like resolved between teams now, everybody's looking at the same data, the same sort of, uh, symptoms and like, Oh yeah, okay. This is telling us, you know, here's the root cause you should investigate this huge, huge thing. Um, and, and it's something we never thought we'd get to where, uh, this, this is where we smart enough to tell us these things, but this, again, this is the power of having all the data under one umbrella >>And the smart analytics. >>Um, and I think really, you know, it's a boat. Uh, if you look at where infrastructure and operations people are today, and especially, you know, eight months, nine months, whatever it is into the pandemic, uh, a lot of them are getting really burnt out with doing the same repetitive tasks over and over again. Um, just trying to keep the lights on, you know, we need, we need to extract those things for those people, uh, just because it just makes no sense to do something over and over again, the same remediation step, just we should automate those things. So getting that sort of, uh, you know, drudgery off their hands, if you will, and, and get them into, into all their important things they should be doing, you know, they're really hard to solve problems. That's where the human shine, um, and that's where, you know, having a, you know, really high level engineers, that's what they should be doing, you know, and just being able to do things I >>Think in a much faster, >>In a more efficient manner, when you think about an incident occurring, right. In, in a level, one technician picks that up and he goes and triaged that maybe run some tests. He has a script, >>Uh, or she, uh, and, >>You know, uh, they open a ticket and they enrich the ticket. They call it some log files. They can look up for the servers on it. You're in an hour and a half into an incident before anyone's even looked at it. If we could automate all of that, >>Why wouldn't we, that makes it easier for everyone. Um, >>Yeah. And I really think that's where the future is, is, is, is bringing this intelligent automation to bear, to take, knock down all the little things that consume the really, the most amount of time. When you think about it, if you aggregate it over the course of a quarter or a year, a great deal of your time is spent just doing that minutiae again, why don't we automate that? And we should. So I really think that's, that's where you get to look long-term. I think also the sense of we're going to be able to measure everything in the sense of business KPIs versus just IT-centric KPIs. That's really where we going to get to in the digital age. And I think we waited too long to do that. I think our operations models were all voted. I think, uh, you know, a lot of, a lot of the KPIs we look at today are completely outmoded. They don't really change if you think about it. When we look at the monthly reports over the course of a year, uh, so let's do something different. And now having all this data and the smart analytics, we can do something different. Absolutely. I'm glad >>That you brought up kind of looking at the impact that AI ops can make on, on minutiae and burnout. That's a really huge problem that so many of us are facing in any industry. And we know that there's some amount of this that's going to continue for a while longer. So let's get our let's leverage intelligent automation to your point, because we can to be able to allow our people to not just be more efficient, but to be making a bigger impact. And there's that mental component there that I think is absolutely critical. I do want to ask you what are some of these? So for those folks going, all right, we've got to do this. It makes sense. We see some short-term things that we need. We need short-term value. We need long-term value as you've just walked us through. What are some of the obstacles that you'd say, Hey, be on the lookout for this to wipe it out of the way. >>Yeah. I, I think there's, you know, when you think about the obstacles, I think people don't think about what are big changes for their organization, right? You know, they're, they're going to change process. They're going to change the way teams interact. They're they're going to change a lot of things, but they're all for the better. So what we're traditionally really bad in infrastructure and operations is communication, marketing, a new initiative, right? We don't go out and get our peers agreement to it where the product owner is, you know, and say, okay, this is what it gets you. This is where it changes. People just hear I'm losing something, I'm losing control over something. You're going to get rid of the tools that I have, but I love I've spent years building out perfecting, um, and that's threatening to people and understandably so because people think if I start losing tools, I start losing head count. >>And then, whereas my department at that point, um, but that's not what this is all about. Uh, this, this isn't a replacement for people. This isn't a replacement for teams. This isn't augmentation. This is getting them back to doing the things they should be doing and less of the stuff they shouldn't be doing. And frankly, it's, it's about providing better services. So when in the end, it's counterintuitive to be against it because it's gonna make it operations look better. It's gonna make us show us that we are the thought leaders in delivering digital services that we can, um, constantly be perfecting the way we're doing it. And Oh, by the way, we can help the business be better. Also at the same time. Uh, I think some of the mistakes people really don't make, uh, really do make, uh, is not looking at their processes today, trying to figure out what they're gonna look like tomorrow when we bring in advanced automation and intelligence, uh, but also being prepared for what the future state is, you know, in talking to one company, they were like, yeah, we're so excited for this. >>Uh, we, we got rid of our old 15 year old laundering system and the same day we stepped a new system. Uh, one problem we had though, was we weren't ready for the amount of incidents that had generated on day one. And it wasn't because we did anything wrong or the system was wrong or what have you. It did the right thing actually, almost too. Well, what it did is it uncovered a lot of really small incidents through advanced correlations. We didn't know we had, so there were things lying out there that were always like, huh, that's weird. That system acts strange sometimes, but we can never pin it down. We found all of those things, which is good. It goes, but it kind of made us all kind of sit back and think, and then our readership are these guys doing their job. Right? >>And then we had to go through an evolution of, you know, just explaining we were 15 years behind from a visibility standpoint to our environment, but technologies that we deployed in applications had moved ahead and modernized. So this is like a cautionary tale of falling too far behind from a sort of a monitoring and intelligence and automation standpoint. Um, so I thought that was a really good story for something like, think about as Eagle would deploy these modern systems. But I think if he really, you know, the marketing to people, so they're not threatened, I think thinking about your process and then what's, what's your day one and then look like, and then what's your six and 12 months after that looks like, I think settling all that stuff upfront just sets you up for success. >>All right. Rich, take us home here. Let's summarize. How can clients build a business case for AI ops? What do you recommend? >>Yeah. You know, I actually get that question a lot. It's usually, uh, almost always the number one, uh, question in, in, um, you know, webinars like this and conversations that, that the audience puts in. So I wouldn't be surprised, but if that was true, uh, going forward from this one, um, yeah, people are like, you know, Hey, we're all in. We want to do this. We know this is the way forward, but the guy who writes the checks, the CIO, the VP of ops is like, you know, I I've signed lots of checks over the years for tools wise is different. Um, and when I guide people to do is to sit back and, and start doing some hard math, right. Uh, one of the things that resonates with the leadership is dollars and cents. It's not percentages. So saying, you know, it's, it brings us a 63% reduction and MTTR is not going to resonate. >>Uh, Oh, even though it's a really good number, you know, uh, I think what it is, you have to put it in terms of avoid, if we could avoid that 63%. Right. You know, um, what does that mean for our, our digital services as far as revenue, right. We know that every hour system down, I think, uh, you know, typically in the market, you see is about $500,000 an hour for enterprise. We'll add that up over the course of the year. What are you losing in revenue? Add to that brand damage loss of customers, you know, uh, Forrester puts out a really big, uh, casino, um, uh, customer experience index every year that measures that if you're delivering good Udall services, bad digital services, if you could raise that up, what does that return to you in revenue? And that's a key thing. And then you just look at the, the, uh, hours of lost productivity. >>I call it, I might call it something else, but I think it's a catchy name. Meaning if a core internal system is down say, and you know, you have a customer service desk of a thousand customer service people, and they can't do that look up or fix that problem for clients for an hour. How much money does that lose you? And you multiply it out. You know, average customer service desk person makes X amount an hour times this much time. This many times it happens. Then you start seeing the real, sort of a power of AI ops for this incident avoidance, or at least lowering the impact of these incidents. And people have put out in graphs and spreadsheets and all this, and then I'm doing some research around this actually to, to, to put out something that people can use to say, the project funds itself in six to 12 months, it's paid for itself. And then after that it's returning money to the business. Why would you not do that? And when you start framing the conversation, that way, the little light bulb turn on for the people that sign the checks. For sure. >>That's great advice for folks to be thinking about. I loved how you talked about the 63% reduction in something. I think that's great. What does it impact? How does it impact the revenue for the organization? If we're avoiding costs here, how do we drive up revenue? So having that laser focus on revenue is great advice for folks in any industry, looking to build a business case for AI ops. I think you set the stage for that rich beautifully, and you were right. This was a fun conversation. Thank you for your time. Thank you. And thanks for watching >>From around the globe with digital coverage. >>Welcome back to the Broadcom AI ops, virtual forum, Lisa Martin here talking with Eastman Nasir global product management at Verizon. We spent welcome back. >>Hi. Hello. Uh, what a pleasure. >>So 2020 the year of that needs no explanation, right? The year of massive challenges and wanting to get your take on the challenges that organizations are facing this year as the demand to deliver digital products and services has never been higher. >>Yeah. So I think this is something it's so close to all the far far, right? It's, uh, it's something that's impacted the whole world equally. And I think regardless of which industry you rent, you have been impacted by this in one form or the other, and the ICT industry, the information and communication technology industry, you know, Verizon being really massive player in that whole arena. It has just been sort of struck with this massive consummation we have talked about for a long time, we have talked about these remote surgery capabilities whereby you've got patients in Kenya who are being treated by an expert sitting in London or New York, and also this whole consciousness about, you know, our carbon footprint and being environmentally conscious. This pandemic has taught us all of that and brought us to the forefront of organization priorities, right? The demand. I think that's, that's a very natural consequence of everybody sitting at home. >>And the only thing that can keep things still going is this data communication, right? But I wouldn't just say that that is, what's kind of at the heart of all of this. Just imagine if we are to realize any of these targets of the world is what leadership is setting for themselves. Hey, we have to be carbon neutral by X year as a country, as a geography, et cetera, et cetera. You know, all of these things require you to have this remote working capabilities, this remote interaction, not just between humans, but machine to machine interactions. And this there's a unique value chain, which is now getting created that you've got people who are communicating with other people or communicating with other machines, but the communication is much more. I wouldn't even use the term real time because we've used real time for voice and video, et cetera. >>We're talking low latency, microsecond decision-making that can either cut somebody's, you know, um, our trees or that could actually go and remove the tumor, that kind of stuff. So that has become a reality. Everybody's asking for it, remote learning, being an extremely massive requirement where, you know, we've had to enable these, uh, these virtual classrooms ensuring the type of connectivity, ensuring the type of type of privacy, which is just so, so critical. You can't just have everybody in a go on the internet and access a data source. You have to be concerned about the integrity and security of that data as the foremost. So I think all of these things, yes, we have not been caught off guard. We were pretty forward-looking in our plans and our evolution, but yes, it's fast track the journey that we would probably believe we would have taken in three years. It has brought that down to two quarters where we've had to execute them. >>Right. Massive acceleration. All right. So you articulated the challenges really well. And a lot of the realities that many of our viewers are facing. Let's talk now about motivations, AI ops as a tool, as a catalyst for helping organizations overcome those challenges. >>So yeah. Now on that I said, you can imagine, you know, it requires microsecond decision-making which human being on this planet can do microsecond decision-making on complex network infrastructure, which is impacting end user applications, which have multitudes of effect. You know, in real life, I use the example of a remote surgeon. Just imagine that, you know, even because of you just use your signal on the quality of that communication for that microsecond, it could be the difference between killing somebody in saving somebody's life. And it's not predictable. We talk about autonomous vehicles. Uh, we talk about this transition to electric vehicles, smart motorways, et cetera, et cetera, in federal environment, how is all of that going to work? You have so many different components coming in. You don't just have a network and security anymore. You have software defined networking. That's coming, becoming a part of that. >>You have mobile edge computing that is rented for the technologies. 5g enables we're talking augmented reality. We're talking virtual reality. All of these things require that resources and why being carbon conscious. We told them we just want to build a billion data centers on this planet, right? We, we have to make sure that resources are given on demand and the best way of resources can be given on demand and could be most efficient is that the thing is being made at million microsecond and those resources are accordingly being distributed, right? If you're relying on people, sipping their coffees, having teas, talking to somebody else, you know, just being away on holiday. I don't think we're going to be able to handle that one that we have already stepped into. Verizon's 5g has already started businesses on that transformational journey where they're talking about end user experience personalization. >>You're going to have events where people are going to go, and it's going to be three-dimensional experiences that are purely customized for you. How, how does that all happen without this intelligence sitting there and a network with all of these multiple layers? So spectrum, it doesn't just need to be intuitive. Hey, this is my private IP traffic. This is public traffic. You know, it has to not be in two, or this is an application that I have to prioritize over another task to be intuitive to the criticality and the context of those transactions. Again, that's surgeons. So be it's much more important than postman setting and playing a video game. >>I'm glad that you think that that's excellent. Let's go into some specific use cases. What are some of the examples that you gave? Let's kind of dig deeper into some of the, what you think are the lowest hanging fruit for organizations kind of pan industry to go after. >>Excellent. Brian, and I think this, this like different ways to look at the lowest hanging fruit, like for somebody like revising who is a managed services provider, you know, very comprehensive medicines, but we obviously have food timing, much lower potentially for some of our customers who want to go on that journey. Right? So for them to just go and try and harness the power of the foods might be a bit higher hanging, but for somebody like us, the immediate ones would be to reduce the number of alarms that are being generated by these overlay services. You've got your basic network, then you've got your whole software defined networking on top of that, you have your hybrid clouds, you have your edge computing coming on top of that. You know? So all of that means if there's an outage on one device on the network, I want to make this very real for everybody, right? >>It's like device and network does not stop all of those multiple applications or monitoring tools from raising and raising thousands of alarm and everyone, one capacity. If people are attending to those thousands of alarms, it's like you having a police force and there's a burglary in one time and the alarm goes off and 50 bags. How, how are you kind of make the best use of your police force? You're going to go investigate 50 bags or do you want to investigate where the problem is? So it's as real as that, I think that's the first wins where people can save so much cost, which is coming from being wasted and resources running around, trying to figure stuff out immediately. I'm tied this with network and security network and security is something which has you did even the most, you know, I mean single screens in our engineering, well, we took it to have network experts, separate people, security experts, separate people to look for different things, but there are security events that can impact the performance of a network. >>And then just drop the case on the side of et cetera, which could be falsely attributed to the metric. And then if you've got multiple parties, which are then the chapter clear stakeholders, you can imagine the blame game that goes on finding fingers, taking names, not taking responsibility that don't has all this happened. This is the only way to bring it all together to say, okay, this is what takes priority. If there's an event that has happened, what is its correlation to the other downstream systems, devices, components, and these are applications. And then subsequently, you know, like isolating it to the right cost where you can most effectively resolve that problem. Thirdly, I would say on demand, virtualized resource, virtualized resources, the heart and soul, the spirit of status that you can have them on demand. So you can automate the allocation of these resources based on customer's consumption their peaks, their cramps, all of that comes in. >>You see, Hey, typically on a Wednesday, the traffic was up significantly for this particular application, you know, going to this particular data center, you could have this automated system, uh, which is just providing those resources, you know, on demand. And so it is to have a much better commercial engagement with customers and just a much better service assurance model. And then one more thing on top of that, which is very critical is that as I was saying, giving that intelligence to the networks to start having context of the criticality of a transaction, that doesn't make sense to them. You can't have that because for that, you need to have this, you know, monkey their data. You need to have multi-cam system, which are monitoring and controlling different aspects of your overall end user application value chain to be communicating with each other. And, you know, that's the only way to sort of achieve that goal. And that only happens with AI. It's not possible >>So it was when you clearly articulated some obvious, low hanging fruit and use cases that organizations can go after. Let's talk now about some of the considerations, you talked about the importance of a network and AI ops, the approach I assume, needs to be modular support needs to be heterogeneous. Talk to us about some of those key considerations that you would recommend. >>Absolutely. So again, basically starting with the network, because if there's, if the metrics sitting at the middle of all of this is not working, then things can communicate with each other, right? And the cloud doesn't work, nothing metal. That's the hardest part of this, but that's the frequency. When you talk about machine to machine communication or IOT, it's just the biggest transformation of the span of every company is going for IOT now to drive those costs, efficiencies, and had, something's got some experience, the integrity of the topic karma, right? The security, integrity of that. How do you maintain integrity of your data beyond just a secure network components? That is true, right? That's where you're getting to the whole arena blockchain technologies, where you have to use digital signatures or barcodes that machine then, and then an intelligence system is automatically able to validate and verify the integrity of the data and the commands that are being executed by those end-user told them what I need to tell them that. >>So it's IOT machines, right? That is paramount. And if anybody is not keeping that into their equation, that in its own self is any system that is therefore maintaining the integrity of your commands and your hold that sits on those, those machines. Right? Second, you have your network. You need to have any else platform, which is able to restless all the fast network information, et cetera. And coupled with that data integrity piece, because for the management, ultimately they need to have a coherent view of the analytics, et cetera, et cetera. They need to know where the problems are again, right? So let's say if there's a problem with the integrity of the commands that are being executed by the machine, that's a much bigger problem than not being able to communicate with that machine and the best thing, because you'd rather not talk to the machine or have to do anything if it's going to start doing wrong things. >>So I think that's where it is. It's very intuitive. It's not true. You have to have subsequently if you have some kind of faith and let me use that use case self autonomous vehicles. Again, I think we're going to see in the next five years, because he's smart with the rates, et cetera, it won't separate autonomous cars. It's much more efficient, it's much more space, et cetera, et cetera. So within that equation, you're going to have systems which will be specialists in looking at aspects and transactions related to those systems. For example, in autonomous moving vehicles, brakes are much more important than the Vipers, right? So this kind of intelligence, it will be multiple systems who have to sit, N nobody has to, one person has to go in one of these systems. I think these systems should be open source enough that they, if you were able to integrate them, right, if something's sitting in the cloud, you were able to integrate for that with obviously the regard of the security and integrity of your data that has to traverse from one system to the other extremely important. >>So I'm going to borrow that integrity theme for a second, as we go into our last question, and that is this kind of take a macro look at the overall business impact that AI ops can help customers make. I'm thinking of, you know, the integrity of teams aligning business in it, which we probably can't talk about enough. We're helping organizations really effectively measure KPIs that deliver that digital experience that all of us demanding consumers expect. What's the overall impact. What would you say in summary fashion? >>So I think the overall impact is a lot of costs. That's customized and businesses gives the time to the time of enterprises. Defense was inevitable. It's something that for the first time, it will come to life. And it's something that is going to, you know, start driving cost efficiencies and consciousness and awareness within their own business, which is obviously going to have, you know, it domino kind of an effect. So one example being that, you know, you have problem isolation. I talked about network security, this multi-layers architecture, which enables this new world of 5g, um, at the heart of all of it, it has to identify the problem to the source, right? Not be bogged down by 15 different things that are going wrong. What is causing those 15 things to go wrong, right? That speed to isolation in its own sense can make millions and millions of dollars to organizations after we organize it. Next one is obviously overall impacted customer experience. Uh, 5g was given out of your customers, expecting experiences from you, even if you're not expecting to deliver them in 2021, 2022, it would have customers asking for those experience or walking away, if you do not provide those experience. So it's almost like a business can do nothing every year. They don't have to reinvest if they just want to die on the line, businesses want remain relevant. >>Businesses want to adopt the latest and greatest in technology, which enables them to, you know, have that superiority and continue it. So from that perspective that continue it, he will read that they write intelligence systems that tank rationalizing information and making decisions supervised by people, of course were previously making some of those. >>That was a great summary because you're right, you know, with how demanding consumers are. We don't get what we want quickly. We churn, right? We go somewhere else and we could find somebody that can meet those expectations. So it has been thanks for doing a great job of clarifying the impact and the value that AI ops can bring to organizations that sounds really now is we're in this even higher demand for digital products and services, which is not going away. It's probably going to only increase it's table stakes for any organization. Thank you so much for joining me today and giving us your thoughts. >>Pleasure. Thank you. We'll be right back with our next segment. >>Digital applications and services are more critical to a positive customer and employee experience than ever before. But the underlying infrastructure that supports these apps and services has become increasingly complex and expanding use of multiple clouds, mobile and microservices, along with modern and legacy infrastructure can make it difficult to pinpoint the root cause when problems occur, it can be even more difficult to determine the business impact your problems that occur and resolve them efficiently. AI ops from Broadcom can help first by providing 360 degree visibility, whether you have hybrid cloud or a cloud native AI ops from Broadcom provides a clear line of sight, including apt to infrastructure and network visibility across hybrid environments. Second, the solution gives you actionable insights by correlating an aggregating data and applying AI and machine learning to identify root causes and even predict problems before users are impacted. Third AI ops from Broadcom provides intelligent automation that identifies potential solutions when problems occur applied to the best one and learns from the effectiveness to improve response in case the problem occurs. Again, finally, the solution enables organizations to achieve digit with jelly by providing feedback loops across development and operations to allow for continuous improvements and innovation through these four capabilities. AI ops from Broadcom can help you reduce service outages, boost, operational efficiency, and effectiveness and improve customer and employee experience. To learn more about AI ops from Broadcom, go to broadcom.com/ai ops from around the globe. >>It's the cube with digital coverage of AI ops virtual forum brought to you by Broadcom. >>Welcome back to the AI ops virtual forum, Lisa Martin here with Srinivasan, Roger Rajagopal, the head of product and strategy at Broadcom. Raj, welcome here, Lisa. I'm excited for our conversation. So I wanted to dive right into a term that we hear all the time, operational excellence, right? We hear it everywhere in marketing, et cetera, but why is it so important to organizations as they head into 2021? And tell us how AI ops as a platform can help. >>Yeah. Well, thank you. First off. I wanna, uh, I want to welcome our viewers back and, uh, I'm very excited to, uh, to share, um, uh, more info on this topic. You know, uh, here's what we believe as we work with large organizations, we see all our organizations are poised to get out of the, uh, the pandemic and look for a brood for their own business and helping customers get through this tough time. So fiscal year 2021, we believe is going to be a combination of, uh, you know, resiliency and agility at the, at the same time. So operational excellence is critical because the business has become more digital, right? There are going to be three things that are going to be more sticky. Uh, you know, remote work is going to be more sticky, um, cost savings and efficiency is going to be an imperative for organizations and the continued acceleration of digital transformation of enterprises at scale is going to be in reality. So when you put all these three things together as a, as a team that is, uh, you know, that's working behind the scenes to help the businesses succeed, operational excellence is going to be, make or break for organizations, >>Right with that said, if we kind of strip it down to the key capabilities, what are those key capabilities that companies need to be looking for in an AI ops solution? >>Yeah, you know, so first and foremost, AI ops means many things to many, many folks. So let's take a moment to simply define it. The way we define AI ops is it's a system of intelligence, human augmented system that brings together full visibility across app infra and network elements that brings together disparate data sources and provides actionable intelligence and uniquely offers intelligent automation. Now, the, the analogy many folks draw is the self-driving car. I mean, we are in the world of Teslas, uh, but you know, uh, but self-driving data center is it's too far away, right? Autonomous systems are still far away. However, uh, you know, application of AI ML techniques to help deal with volume velocity, veracity of information, uh, is, is critical. So that's how we look at AI ops and some of the key capabilities that we, uh, that we, uh, that we work with our customers to help them on our own for eight years. >>Right? First one is eyes and ears. What we call full stack observability. If you do not know what is happening in your systems, uh, you know, that that serve up your business services. It's going to be pretty hard to do anything, uh, in terms of responsiveness, right? So from stack observability, the second piece is what we call actionable insights. So when you have disparate data sources, tools, sprawls data coming at you from, uh, you know, uh, from a database systems, it systems customer management systems, ticketing systems. How do you find the needle from the haystack? And how do you respond rapidly from a myriad of problems as CEO of red? The third area is what we call intelligent automation. Well, identifying the problem to act on is important, and then acting on automating that and creating, uh, a recommendation system where, uh, you know, you can be proactive about it is even more important. And finally, all of this focuses on efficiency. What about effectiveness? Effectiveness comes when you create a feedback loop, when what happens in production is related to your support systems and your developers so that they can respond rapidly. So we call that continuous feedback. So these are the four key capabilities that, uh, you know, uh, you should look for in an AI ops system. And that's what we offer as well. >>Russia, there's four key capabilities that businesses need to be looking for. I'm wondering how those help to align business. And it it's, again like operational excellence. It's something that we talk about a lot is the alignment of business. And it a lot more challenging, easier said than done, right. But I want you to explain how can AI ops help with that alignment and align it outputs to business outcomes? >>Yeah. So, you know, one of the things, uh, I'm going to say something that is, uh, that is, uh, that is simple, but, but, but this harder, but alignment is not on systems alignment is with people, right? So when people align, when organizations align, when cultures align, uh, dramatic things can happen. So in the context of AI ops VC, when, when SRE is aligned with the DevOps engineers and information architects and, uh, uh, you know, it operators, uh, you know, they enable organizations to reduce the gap between intent and outcome or output and outcome that said, uh, you know, these personas need mechanisms to help them better align, right. Help them better visualize, see the, you know, what we call single source of truth, right? So there are four key things that I want to call out. When we work with large enterprises, we find that customer journey alignment with the, you know, what we call it systems is critical. >>So how do you understand your business imperatives and your customer journey goals, whether it is car to a purchase or whether it is, uh, you know, bill shock scenarios and Swan alignment on customer journey to your it systems is one area that you can reduce the gap. The second area is how do you create a scenario where your teams can find problems before your customers do right outage scenarios and so on. So that's the second area of alignment. The third area of alignment is how can you measure business impact driven services? Right? There are several services that an organization offers versus an it system. Some services are more critical to the business than others, and these change in a dynamic environment. So how do you, how do you understand that? How do you measure that and how, how do you find the gaps there? So that's the third area of alignment that we, that we help and last but not least there are, there are things like NPS scores and others that, that help us understand alignment, but those are more long-term. But in the, in the context of, uh, you know, operating digitally, uh, you want to use customer experience and business, uh, you know, a single business outcome, uh, as a, as a key alignment factor, and then work with your systems of engagement and systems of interaction, along with your key personas to create that alignment. It's a people process technology challenge. >>So, whereas one of the things that you said there is that it's imperative for the business to find a problem before a customer does, and you talked about outages there, that's always a goal for businesses, right. To prevent those outages, how can AI ops help with that? Yeah, >>So, you know, outages, uh, talk, you know, go to resiliency of a system, right? And they also go to, uh, uh, agility of the same system, you know, if you're a customer and if you're whipping up your mobile app and it takes more than three milliseconds, uh, you know, you're probably losing that customer, right. So outages mean different things, you know, and there's an interesting website called down detector.com that actually tracks all the old pages of publicly available services, whether it's your bank or your, uh, you know, tele telecom service or a mobile service and so on and so forth. In fact, the key question around outages for, from, uh, from, uh, you know, executives are the question of, are you ready? Right? Are you ready to respond to the needs of your customers and your business? Are you ready to rapidly resolve an issue that is impacting customer experience and therefore satisfaction? >>Are you creating a digital trust system where customers can be, you know, um, uh, you know, customers can feel that their information is secure when they transact with you, all of these, getting into the notion of resiliency and outages. Now, you know, one of the things that, uh, that I, I often, uh, you know, work with customers around, you know, would that be find as the radius of impact is important when you deal with outages? What I mean by that is problems occur, right? How do you respond? How quickly do you take two seconds, two minutes, 20 minutes, two hours, 20 hours, right? To resolve the problem that radius of impact is important. That's where, you know, you have to bring a gain people, process technology together to solve that. And the key thing is you need a system of intelligence that can aid your teams, you know, look at the same set of parameters so that you can respond faster. That's the key here. >>We look at digital transformation at scale. Raj, how does AI ops help influence that? >>You know, um, I'm going to take a slightly long-winded way to answer this question. See when it comes to digital transformation at scale, the focus on business purpose and business outcome becomes extremely critical. And then the alignment of that to your digital supply chain, right, are the, are the, are the key factors that differentiate winners in the, in their digital transformation game? Really, what we have seen, uh, with, with winners is they operate very differently. Like for example, uh, you know, Nike matures, its digital business outcomes by shoes per second, right? Uh, Apple by I-phones per minute, Tesla by model threes per month, are you getting this, getting it right? I mean, you want to have a clear business outcome, which is a measure of your business, uh, in effect, I mean, ENC, right? Which, which, uh, um, my daughter use and I use very well. >>Right. Uh, you know, uh, they measure by revenue per hour, right? I mean, so these are key measures. And when you have a key business outcome measure like that, you can everything else, because you know what these measures, uh, you know, uh, for a bank, it may be deposits per month, right now, when you move money from checking account to savings account, or when you do direct deposits, those are, you know, banks need liquidity and so on and so forth. But, you know, the, the key thing is that single business outcome has a Starburst effect inside the it organization that touches a single money moment from checking a call to savings account can touch about 75 disparate systems internally. Right? So those think about it, right? I mean, all, all we're doing is moving money from checking account a savings account. Now that goats into a it production system, there are several applications. >>There is a database, there is, there are infrastructures, there are load balancers that are webs. You know, you know, the web server components, which then touches your, your middleware component, which is a queuing system, right. Which then touches your transactional system. Uh, and, uh, you know, which may be on your main frames, what we call mobile to mainframe scenario, right? And we are not done yet. Then you have a security and regulatory compliance system that you have to touch a fraud prevention system that you have to touch, right? A state department regulation that you may have to meet and on and on and on, right? This is the chat that it operations teams face. And when you have millions of customers transacting, right, suddenly this challenge cannot be managed by human beings alone. So therefore you need a system of intelligence that augments human intelligence and acts as your, you know, your, your eyes and ears in a way to, to point pinpoint where problems are. >>Right. So digital transformation at scale really requires a very well thought out AI ops system, a platform, an open extensible platform that, uh, you know, uh, that is heterogeneous in nature because there's tools, products in organizations. There is a lot of databases in systems. There are millions of, uh, uh, you know, customers and hundreds of partners and vendors, you know, making up that digital supply chain. So, you know, AI ops is at the center of an enabling an organization achieve digital op you know, transformation at scale last but not least. You need continuous feedback loop. Continuous feedback loop is the ability for a production system to inform your dev ops teams, your finance teams, your customer experience teams, your cost modeling teams about what is going on so that they can so that they can reduce the intent, come gap. >>All of this need to come together, what we call BizOps. >>That was a great example of how you talked about the Starburst effect. I actually never thought about it in that way, when you give the banking example, but what you should is the magnitude of systems. The fact that people alone really need help with that, and why intelligent automation and AI ops can be transformative and enable that scale. Raj, it's always a pleasure to talk with you. Thanks for joining me today. And we'll be right back with our next segment. Welcome back to the AI ops virtual forum. We've heard from our guests about the value of AI ops and why and how organizations are adopting AI ops platforms. But now let's see AI ops inaction and get a practical view of AI ops to deep Dante. The head of AI ops at Broadcom is now going to take you through a quick demo. >>Hello. So they've gotta head off AI ops and automation here. What I'm going to do today is talk through some of the key capabilities and differentiators of Broadcom's CII ops solution in this solution, which can be delivered on cloud or on-prem. We bring a variety of metric alarm log and applauded data from multiple sources, EPM, NetApps, and infrastructure monitoring tools to provide a single point of observability and control. Let me start where our users mostly stock key enterprises like FSI, telcos retailers, et cetera, do not manage infrastructure or applications without having a business context. At the end of the day, they offer business services governed by SLS service level objectives and SLI service level indicators are service analytics, which can scale to a few thousand services, lets our customers create and monitor the services as per their preference. They can create a hierarchy of services based on their business practice. >>For example, here, the sub services are created based on functional subsistence for certain enterprises. It could be based on location. Users can import these services from their favorite CMDB. What's important to note that not all services are born equal. If you are a modern bank, you may want to prioritize tickets coming from digital banking, for example, and this application lets you rank them as per the KPI of your choice. We can source the availability, not merely from the state of the infrastructure, whether they're running or not. But from the SLS that represent the state of the application, when it comes to triaging issues related to the service, it is important to have a complete view of the topology. The typology can show both east-west elements from mobile to mainframe or not South elements in a network flow. This is particularly relevant for a large enterprise who could be running the systems of engagement on the cloud and system of records on mainframe inside the firewall here, you can see that the issue is related to the mainframe kick server. >>You can expand to see the actual alarm, which is sourced from the mainframe operational intelligence. Similarly, clicking on network will give the hub and spoke view of the network devices, the Cisco switches and routers. I can click on the effected router and see all the details Broadcom's solution stores, the ontological model of the typology in the form of a journal graph where one can not only view the current state of the typology, but the past as well, talking of underlying data sources, the solution uses best of the pre data stores for structured and unstructured data. We have not only leveraged the power of open source, but have actively contributed back to the community. One of the key innovations is evident in our dashboarding framework because we have enhanced the open source Grafana technology to support these diverse data sources here. You can see a single dashboard representing applications to infrastructure, to mainframe again, sourcing a variety of data from these sources. >>When we talk to customers, one of the biggest challenges that they face today is related to alarms because of a proliferation of tools. They are currently drowning in an ocean of hundreds and thousands of alarms. This increases the Elmont support cost to tens of dollars per ticket, and also affects LTO efficiency leading to an average of five to six hours of meantime to resolution here is where we have the state of the art innovation utilizing the power of machine learning and ontology to arrive at the root cause we not only clusterize alarms based on text, but employ the technique of 41st. We look at the topology then at the time window duplicate text based on NLP. And lastly learn from continuous training of the model to deduce what we call situations. This is an example of a situation. As you can see, we provide a time-based evidence of how things unfolded and arrive at a root cause. >>Lastly, the solution provides a three 60 degree closed loop remediation either through a ticketing system or by direct invocation of automation actions instead of firing hard-coded automation runbooks for certain conditions, the tool leverage is machine learning to rank automation actions based on past heuristics. That's why we call it intelligent automation to summarize AI ops from Broadcom helps you achieve operational excellence through full stack observability, coupled with AIML that applies across modern hybrid cloud environments, as well as legacy ones uniquely. It ties these insights with intelligent automation to improve customer experience. Thank you for watching from around the globe. It's the cube with digital coverage of AI ops virtual forum brought to you by Broadcom. >>Welcome to our final segment today. So we've discussed today. The value that AI ops will bring to organizations in 2021, we'll discuss that through three different perspectives. And so now we want to bring those perspectives together and see if we can get a consensus on where AI ops needs to go for folks to be successful with it in the future. So bringing back some folks Richland is back with us. Senior analysts, serving infrastructure and operations professionals at Forrester smartness here is also back in global product management at Verizon and Srinivasan, Reggie Gopaul head of product and strategy at Broadcom guys. Great to have you back. So let's jump in and rich, we're going to, we're going to start with you, but we are going to get all three of you, a chance to answer the questions. So we've talked about why organizations should adopt AI ops, but what happens if they choose not to what challenges would they face? Basically what's the cost of organizations doing nothing >>Good question, because I think in operations for a number of years, we've kind of stand stood, Pat, where we are, where we're afraid change things sometimes, or we just don't think about a tooling as often. The last thing to change because we're spending so much time doing project work and modernization and fighting fires on a daily basis. >>Problem is going to get worse. If we do nothing, >>You know, we're building new architectures like containers and microservices, which means more things to mind and keep running. Um, we're building highly distributed systems. We're moving more and more into this hybrid world, a multi-cloud world, uh, it's become over-complicate and I'll give a short anecdote. I think, eliminate this. Um, when I go to conferences and give speeches, it's all infrastructure operations people. And I say, you know, how many people have three X, five X, you know, uh, things to monitor them. They had, you know, three years ago, two years ago, and everyone's saying how many people have hired more staff in that time period, zero hands go up. That's the gap we have to fill. And we have to fill that through better automation, more intelligent systems. It's the only way we're going to be able to fill back out. >>What's your perspective, uh, if organizations choose not to adopt AI ops. Yeah. So I'll do that. Yeah. So I think it's, I would just relate it to a couple of things that probably everybody >>Tired off lately and everybody can relate to. And this would resonate that we have 5g, which is all set to transform the world. As we know it, I don't have a lot of communication with these smart cities, smart communities, IOT, which is going to make us pivotal to the success of businesses. And as you've seen with this call with, you know, transformation of the world, that there's a, there's a much bigger cost consciousness out there. People are trying to become much more, forward-looking much more sustainable. And I think at the heart of all of this, that the necessity that you have intelligent systems, which are bastardizing more than enough information that previously could've been overlooked because if you don't measure engagement, not going right. People not being on the same page of this using two examples or hundreds of things, you know, that play a part in things, but not coming together in the best possible way. So I think it has an absolute necessity to drive those cost efficiencies rather than, you know, left right and center laying off people who are like 10 Mattel to your business and have a great tribal knowledge of your business. So to speak, you can drive these efficiencies through automating a lot of those tasks that previously were being very manually intensive or resource intensive. And you could allocate those resources towards doing much better things, which let's be very honest going into 20, 21 after what we've seen with 2020, it's going to be mandate treat. >>And so Raj, I saw you shaking your head there when he was mom was sharing his thoughts. What are your thoughts about that sounds like you agree. Yeah. I mean, uh, you know, uh, to put things in perspective, right? I mean we're firmly in the digital economy, right? Digital economy, according to the Bureau of economic analysis is 9% of the U S GDP. Just, you know, think about it in, in, in, in, in the context of the GDP, right? It's only ranked lower, slightly lower than manufacturing, which is at 11.3% GDP and slightly about finance and insurance, which is about seven and a half percent GDP. So the digital economy is firmly in our lives, right. And as Huisman was talking about it, you know, software eats the world and digital, operational excellence is critical for customers, uh, to, uh, you know, to, uh, to drive profitability and growth, uh, in the digital economy. >>It's almost, you know, the key is digital at scale. So when, uh, when rich talks about some of the challenges and when Huseman highlights 5g as an example, those are the things that, that, that come to mind. So to me, what is the cost or perils of doing nothing? You know, uh, it's not an option. I think, you know, more often than not, uh, you know, C-level execs are asking head of it and they are key influencers, a single question, are you ready? Are you ready in the context of addressing spikes in networks because of the pandemic scenario, are you ready in the context of automating away toil? Are you ready to respond rapidly to the needs of the digital business? I think AI ops is critical. >>That's a great point. Roger, where does stick with you? So we got kind of consensus there, as you said, wrapping it up. This is basically a, not an option. This is a must to go forward for organizations to be successful. So let's talk about some quick wins, or as you talked about, you know, organizations and sea levels asking, are you ready? What are some quick wins that that organizations can achieve when they're adopting AI? >>You know, um, immediate value. I think I would start with a question. How often do your customers find problems in your digital experience before you do think about that? Right. You know, if you, if you, you know, there's an interesting web, uh, website, um, uh, you know, down detector.com, right? I think, uh, in, in Europe there is an equal amount of that as well. It ha you know, people post their digital services that are down, whether it's a bank that, uh, you know, customers are trying to move money from checking account, the savings account and the digital services are down and so on and so forth. So some and many times customers tend to find problems before it operations teams do. So a quick win is to be proactive and immediate value is visibility. If you do not know what is happening in your complex systems that make up your digital supply chain, it's going to be hard to be responsive. So I would start there >>Visibility this same question over to you from Verizon's perspective, quick wins. >>Yeah. So I think first of all, there's a need to ingest this multi-care spectrum data, which I don't think is humanly possible. You don't have people having expertise, you know, all the seven layers of the OSI model and then across network and security and at the application level. So I think you need systems which are now able to get that data. It shouldn't just be wasted reports that you're paying for on a monthly basis. It's about time that you started making the most of those in the form of identifying what are the efficiencies within your ecosystem. First of all, what are the things, you know, which could be better utilized subsequently you have the >>Opportunity to reduce the noise of a trouble tickets handling. It sounds pretty trivial, but >>An average you can imagine every trouble tickets has the cost in dollars, right? >>So, and there's so many tickets and there's art >>That get created on a network and across an end user application value, >>We're talking thousands, you know, across and end user >>Application value chain could be million in >>A year. So, and so many of those are not really, >>He, you know, a cause of concern because the problem is something. >>So I think that whole triage is an immediate cost saving and the bigger your network, the bigger >>There's a cost of things, whether you're a provider, whether you're, you know, the end customer at the end of the day, not having to deal with problems, which nobody can resolve, which are not meant to be dealt with. There's so many of those situations, right, where service has just been adopted, >>Which is just coordinate quality, et cetera, et cetera. So many reasons. So those are the, >>So there's some of the immediate cost saving them. They are really, really significant. >>Secondly, I would say Raj mentioned something about, you know, the user, >>Your application value chain, and an understanding of that, especially with this hybrid cloud environment, >>Et cetera, et cetera, right? The time it takes to identify a problem in an end user application value chain across the seven layers that I mentioned with the OSI reference model across network and security and the application environment. It's something that >>In its own self has massive cost to business, >>Right? That could be >>No sale transactions that could be obstructed because of this. There could be, and I'm going to use a really interesting example. >>We talk about IOT. The integrity of the IOT machine is exciting. >>Family is pivotal in this new world that we're stepping into. >>You could be running commands, >>Super efficient. He has, everything is being told to the machine really fast with sending yeah. >>Everything there. What if it's hacked? And if that's okay, >>Robotic arm starts to involve the things you don't want it to do. >>So there's so much of that. That becomes a part of this naturally. And I believe, yes, this is not just like from a cost >>standpoint, but anything going wrong with that code base, et cetera, et cetera. These are massive costs to the business in the form of the revenue. They have lost the perception in the market as a result, the fed, >>You know, all that stuff. So >>These are a couple of very immediate problems, but then you also have the whole player virtualized resources where you can automate the allocation, you know, the quantification of an orchestration of those virtualized resources, rather than a person having to, you know, see something and then say, Oh yeah, I need to increase capacity over here, because then it's going to have this particular application. You have systems doing this stuff and to, you know, Roger's point your customer should not be identifying your problems before you, because this digital is where it's all about perception. >>Absolutely. We definitely don't want the customers finding it before. So rich, let's wrap this particular question up with you from that senior analyst perspective, how can companies use make big impact quickly with AI ops? Yeah, >>Yeah, I think, you know, and it was been really summed up some really great use cases there. I think with the, uh, you know, one of the biggest struggles we've always had in operations is isn't, you know, the mean time to resolve. We're pretty good at resolving the things. We just have to find the thing we have to resolve. That's always been the problem and using these advanced analytics and machine learning algorithms now across all machine and application data, our tendency is humans is to look at the console and say, what's flashing red. That must be what we have to fix, but it could be something that's yellow, somewhere else, six services away. And we have made things so complicated. And I think this is what it was when I was saying that we can't get there anymore on our own. We need help to get there in all of this stuff that the outline. >>So, so well builds up to a higher level thing of what is the customer experience about what is the customer journey? And we've struggled for years in the digital world and measuring that a day-to-day thing. We know an online retail. If you're having a bad experience at one retailer, you just want your thing. You're going to go to another retailer, brand loyalty. Isn't one of like it, wasn't a brick and mortal world where you had a department store near you. So you were loyal to that because it was in your neighborhood, um, online that doesn't exist anymore. So we need to be able to understand the customer from that first moment, they touch a digital service all the way from their, their journey through that digital service, the lowest layer, whether it be a database or the network, what have you, and then back to them again, and we're not understanding, is that a good experience? >>We gave them. How does that compare to last week's experience? What should we be doing to improve that next week? Uh, and I think companies are starting and then the pandemic certainly, you know, push this timeline. If you listened to the, the, the CEO of Microsoft, he's like, you know, 10 years of digital transformation written down. And the first several months of this, um, in banks and in financial institutions, I talked to insurance companies, aren't slowing down. They're trying to speed up. In fact, what they've discovered is that they're, you know, obviously when we were on lockdown or what have you, they use of digital servers is spiked very high. What they've learned is they're never going to go back down. They're never going to return to pretend endemic levels. So now they're stuck with this new reality. Well, how do we service those customers and how do we make sure we keep them loyal to our brand? >>Uh, so, you know, they're looking for modernization opportunities. A lot of that that's things have been exposed. And I think Raj touched upon this very early in the conversation is visibility gaps. Now that we're on the outside, looking in at the data center, we know we architect things in a very way. Uh, we better ways of making these correlations across the Sparrow technologies to understand where the problems lies. We can give better services to our customers. And I think that's really what we're going to see a lot of the innovation and the people really clamoring for these new ways of doing things that starting, you know, now, I mean, I've seen it in customers, but I think really the push through the end of this year to next year when, you know, economy and things like that straightened out a little bit more, I think it really, people are gonna take a hard look of where they are and is, you know, AI ops the way forward for them. And I think they'll find it. The answer is yes, for sure. >>So we've, we've come to a consensus that, of what the parallels are of organizations, basically the cost of doing nothing. You guys have given some great advice on where some of those quick wins are. Let's talk about something Raj touched on earlier is organizations, are they really ready for truly automated AI? Raj, I want to start with you readiness factor. What are your thoughts? >>Uh, you know, uh, I think so, you know, we place our, her lives on automated systems all the time, right? In our, in our day-to-day lives, in the, in the digital world. I think, uh, you know, our, uh, at least the customers that I talk to our customers are, uh, are, uh, you know, uh, have a sophisticated systems. Like for example, advanced automation is a reality. If you look at social media, AI and ML and automation are used to automate away, uh, misinformation, right? If you look at financial institutions, AI and ML are used to automate away a fraud, right? So I want to ask our customers why can't we automate await oil in it, operation systems, right? And that's where our customers are. Then the, you know, uh, I'm a glass half full, uh, cleanup person, right? Uh, this pandemic has been harder on many of our customers, but I think what we have learned from our customers is they've Rose to the occasion. >>They've used digital as a key needs, right? At scale. That's what we see with, you know, when, when Huseman and his team talk about, uh, you know, network operational intelligence, right. That's what it means to us. So I think they are ready, the intersection of customer experience it and OT, operational technology is ripe for automation. Uh, and, uh, you know, I, I wanna, I wanna sort of give a shout out to three key personas in this mix. It's about people, right? One is the SRE persona, you know, site, reliability engineer. The other is the information security persona. And the third one is the it operator automation engineer persona. These folks in organizations are building a system of intelligence that can respond rapidly to the needs of their digital business. We at Broadcom, we are in the business of helping them construct a system of intelligence that will create a human augmented solution for them. Right. So when I see, when I interact with large enterprise customers, I think they, they, you know, they, they want to achieve what I would call advanced automation and AI ML solutions. And that's squarely, very I ops is, you know, is going as it, you know, when I talk to rich and what, everything that rich says, you know, that's where it's going and that's what we want to help our customers to. So, which about your perspective of organizations being ready for truly automated AI? >>I think, you know, the conversation has shifted a lot in the last, in, in pre pandemic. Uh, I'd say at the end of last year, we're, you know, two years ago, people I'd go to conferences and people come up and ask me like, this is all smoke and mirrors, right? These systems can't do this because it is such a leap forward for them, for where they are today. Right. We we've sort of, you know, in software and other systems, we iterate and we move forward slowly. So it's not a big shock. And this is for a lot of organizations that big, big leap forward where they're, they're running their operations teams today. Um, but now they've come around and say, you know what? We want to do this. We want all the automations. We want my staff not doing the low complexity, repetitive tasks over and over again. >>Um, you know, and we have a lot of those kinds of legacy systems. We're not going to rebuild. Um, but they need certain care and feeding. So why are we having operations? People do those tasks? Why aren't we automating those out? I think the other piece is, and I'll, I'll, I'll send this out to any of the operations teams that are thinking about going down this path is that you have to understand that the operations models that we're operating under in, in INO and have been for the last 25 years are super outdated and they're fundamentally broken for the digital age. We have to start thinking about different ways of doing things and how do we do that? Well, it's, it's people, organization, people are going to work together differently in an AI ops world, um, for the better. Um, but you know, there's going to be the, the age of the 40 person bridge call thing. >>Troubleshooting is going away. It's going to be three, four, five focused engineers that need to be there for that particular incident. Um, a lot of process mailer process we have in our level, one level, two engineering. What have you running of tickets, gathering of artifacts, uh, during an incident is going to be automated. That's a good thing. We should be doing those, those things by hand anymore. So I'd say that the, to people's like start thinking about what this means to your organization. Start thinking about the great things we can do by automating things away from people, having to do them over and over again. And what that means for them, getting them matched to what they want to be doing is high level engineering tasks. They want to be doing monitorization, working with new tools and technologies. Um, these are all good things that help the organization perform better as a whole great advice and great kind of some of the thoughts that you shared rich for what the audience needs to be on the lookout. For one, I want to go over to you, give me your thoughts on what the audience that should be on the lookout for, or put on your agendas in the next 12 months. >>So there's like a couple of ways to answer that question. One thing would be in the form of, you know, what are some of the things they have to be concerned about in terms of implementing this solution or harnessing its power. The other one could be, you know, what are the perhaps advantages they should look to see? So if I was to talk about the first one, let's say that, what are some of the things I have to watch out for like possible pitfalls that everybody has data, right? So yeah, there's one strategy we say, okay, you've got the data, let's see what we can do with them. But then there's the exact opposite side, which has to be considered when you're doing that analysis. What are the use cases that you're looking to drive? Right. But then use cases you have to understand, are you taking a reactive use case approach? >>Are you taking active use cases, right? Or, yeah, that's a very, very important concentration. Then you have to be very cognizant of where does this data that you have, where does it reside? What are the systems and where does it need to go to in order for this AI function to happen and subsequently if there needs to be any backward communication with all of that data in a process manner. So I think these are some of the very critical points because you can have an AI solution, which is sitting in a customer data center. It could be in a managed services provider data center, like, right, right. It could be in a cloud data center, like an AWS or something, or you could have hybrid views, et cetera, all of that stuff. So you have to be very mindful of where you're going to get the data from is going to go to what are the use cases you're trying to get out to do a bit of backward forward. >>Okay, we've got this data thing and I think it's a journey. Nobody can come in and say, Hey, you've built this fantastic thing. It's like Terminator two. I think it's a journey where we built starting with the network. My personal focus always comes down to the network and with 5g so much, so much more right with 5g, you're talking low latency communication. That's like the true power of 5g, right? It's low latency, it's ultra high bandwidth, but what's the point of that low latency. If then subsequently the actions that need to be taken to prevent any problems in application, IOT applications, remote surgeries, uh, self driving vehicles, et cetera, et cetera. What if that's where people are sitting and sipping their coffees and trying to take action that needs to be in low latency as well. Right? So these are, I think some of the fundamental things that you have to know your data, your use cases, that location, where it needs to be exchanged, what are the parameters around that for extending that data? >>And I think from that point at one word, it's all about realizing, you know, sense of business outcomes. Unless AI comes in as a digital labor that shows you, I have, I have reduced your this amount of time and that's a result of big problems or identified problems for anything. Or I have saved you this much resource in a month, in a year or whatever timeline that people want to see it. So I think those are some of the initial starting points, and then it all starts coming together. But the key is it's not one system that can do everything. You have to have a way where, you know, you can share data once you've caught all of that data into one system. Maybe you can send it to another system at make more, take more advantage, right? That system might be an AI and IOT system, which is just looking at all of your street and make it sure that Hey parents. So it's still off just to be more carbon neutral and all that great stuff, et cetera, et cetera, >>Stuff for the audience to can cigarette rush, take us time from here. What are some of the takeaways that you think the audience really needs to be laser focused on as we move forward into the next year? You know, one thing that, uh, I think a key takeaway is, um, uh, you know, as we embark on 2021, closing the gap between intent and outcome and outputs and outcome will become critical, is critical. Uh, you know, especially for, uh, you know, uh, digital transformation at scale for organizations context in the, you know, for customer experience becomes even more critical as who Swan Huseman was talking, uh, you know, being network network aware network availability is, is a necessary condition, but not sufficient condition anymore. Right? The what, what, what customers have to go towards is going from network availability to network agility with high security, uh, what we call app aware networks, right? How do you differentiate between a trade, a million dollar trade that's happening between, uh, you know, London and New York, uh, uh, versus a YouTube video training that an employee is going through? Worse is a YouTube video that millions of customers are, are >>Watching, right? Three different context, three different customer scenarios, right? That is going to be critical. And last but not least feedback loop, uh, you know, responsiveness is all about feedback loop. You cannot predict everything, but you can respond to things faster. I think these are sort of the three, three things that, uh, that, uh, you know, customers aren't going to have to have to really think about. And that's also where I believe AI ops, by the way, AI ops and I I'm. Yeah. You know, one of the points that was smart and shout out to what he was saying was heterogeneity is key, right? There is no homogeneous tool in the world that can solve problems. So you want an open extensible system of intelligence that, that can harness data from disparate data sources provide that visualization, the actionable insight and the human augmented recommendation systems that are so needed for, uh, you know, it operators to be successful. I think that's where it's going. >>Amazing. You guys just provided so much content context recommendations for the audience. I think we accomplished our goal on this. I'll call it power panel of not only getting to a consensus of what, where AI ops needs to go in the future, but great recommendations for what businesses in any industry need to be on the lookout for rich Huisman Raj, thank you for joining me today. We want to thank you for watching. This was such a rich session. You probably want to watch it again. Thanks for your time. Thanks so much for attending and participating in the AI OBS virtual forum. We really appreciate your time and we hope you really clearly understand the value that AI ops platforms can deliver to many types of organizations. I'm Lisa Martin, and I want to thank our speakers today for joining. We have rich lane from Forrester who's fund here from Verizon and Raj from Broadcom. Thanks everyone. Stay safe..

>>From around the globe with digital coverage brought to you by Broadcom. >>So our final segment today, so we've discussed today, the value that AI ops will bring to organizations in 2021, we'll discuss that through three different perspectives. And so now we want to bring those perspectives together and see if we can get a consensus on where AI ops needs to go for folks to be successful with it in the future. So bringing back some folks Richland is back with us. Senior analysts, serving infrastructure and operations professionals at Forester with smartness here is also back global product management at Verizon and Srinivasan, Reggie Gopaul head of product and strategy at Broadcom guys. Great to have you back. So let's jump in and Richard, we're going to, we're going to start with you, but we are going to get all three of you, a chance to answer the questions. So rich, we've talked about why organizations should adopt AI ops, but what happens if they choose not to what challenges would they face? Basically what's the cost of organizations doing nothing. >>Yeah. So it's a really good question because I think in operations for a number of reviews, we've kind of stand, uh, stood Pat, where we are, where we're afraid change things sometimes. Or we just don't think about a tooling is often the last thing to change because we're spending so much time doing project work and modernization and fighting fires on a daily basis. Uh, that problem is going to get worse if we do nothing. Um, you know, we're building new architectures like containers and microservices, which means more things to mind and keep running. Um, we're building highly distributed systems where you got moving more and more into this hybrid world, the multicloud world, uh, it's become over-complicate and I'll give a short anecdote. I think, eliminate this. Um, when I go to conferences and give speeches, it's all infrastructure operations people. And I say, you know, how many people have three X, five X, you know, uh, things to monitor them. They had, you know, three years ago, two years ago, and everyone's hand goes up, say how many people have hired more staff in that time period. Zero hands go up. That's the gap we have to fill. And we have to fill that through better automation, more intelligent systems. It's the only way we're going to be able to feel back out. >>What's your perspective, uh, if organizations choose not to adopt AI ops. >>Yeah. >>That's pretty good. So I'll do that. >>Yeah. So I think it said, I say it's related to a couple of things that probably everybody tired off lately and everybody can relate to. And this would resonate that we'd have 5g, which is old set to transform the one that we know it, of communication with these smart cities, smart communities, IOT, which is going to become pivotal to the success of businesses. And as we seen with this, COVID, you know, transformation of the world that there's a, there's a much bigger cost consciousness out there. People are trying to become much more, forward-looking much more sustainable. And I think at the heart of all of this, that the necessity that you have intelligent systems, which are bastardizing more than enough information that previous equipment overlooked, because if you don't measure engagement, not going right. People love being on the same page of this using two examples for hundreds of things that play a part in things not coming together in the best possible way. So I think it has an absolute necessity to grind those cost efficiencies rather than, you know, left right and center laying off people who are like pit Mattel to your business and have a great tribal knowledge of your business. So to speak, you can drive these efficiencies through automating a lot of those tasks that previously were being very manually intensive or resource intensive. And you could allocate those resources towards doing much better things, which let's be very honest going into 20, 21, after what we've seen with 2020, it's going to be mandate >>Shaking your head there when you, his mom was sharing his thoughts. What are your thoughts about this sounds like you agree. Yeah. I mean, uh, you know, uh, to put things in perspective, right? I mean, we are firmly in the digital economy, right? Digital economy, according to the Bureau of economic analysis is 9% of the us GDP. Just, you know, think about it in, in, in, in, in the context of the GDP, right? It's only ranked lower, slightly lower than manufacturing, which is at 11.3% GDP and slightly about finance and insurance, which is about seven and a half percent GDP. So G the digital economy is firmly in our lives, right? And so someone was talking about it, you know, software eats the world and digital, operational excellence is critical for customers, uh, to, uh, you know, to, uh, to drive profitability and growth, uh, in the digital economy. >>It's almost, you know, the key is digital at scale. So when, uh, when rich talks about some of the challenges and when newsman highlights 5g, as an example, those are the things that, that, that come to mind. So to me, what is the cost or perils of doing nothing? You know, uh, it's not an option. I think, you know, more often than not, uh, you know, C-level execs are asking their head of it and they are key influencers, a single question, are you ready? Are you ready in the context of addressing spikes in networks because of, uh, the pandemic scenario, are you ready in the context of automating away toil? Are you ready to respond rapidly to the needs of the digital business? I think AI ops is critical. >>That's a great point, Roger, where gonna stick with you. So we got kind of consensus there, as you said, wrapping it up. This is basically a, not an option. This is a must to go forward for organizations to be successful. So let's talk about some quick wins. So as you talked about, you know, organizations and C-levels asking, are you ready? What are some quick wins that that organizations can achieve when they're adopting AI? >>You know, um, immediate value. I think I would start with a question. How often do your customers find problems in your digital experience before you think about that? Right. You know, if you, if you, you know, there's an interesting web, uh, website, um, uh, you know, down detector.com, right? I think, uh, in, in Europe, there is an equal amount of that as well. It ha you know, people post their digital services that are down, whether it's a bank that, uh, you know, customers are trying to move money from checking account, the savings account and the digital services are down and so on and so forth. So some and many times customers tend to find problems before it operation teams do. So. A quick win is to be proactive and immediate value is visibility. If you do not know what is happening in your complex systems that make up your digital supply chain, it's going to be hard to be responsive. So I would start there >>Vice visibility. There's some question over to you from Verizon's perspective, quick wins. >>Yeah. So I think first of all, there's a need to ingest this multi-layered monetize spectrum data, which I don't think is humanly possible. You don't have people having expertise, you know, all seven layers of the OSI model and then across network and security and at the application of it. So I think you need systems which are now able to get that data. It shouldn't just be wasted reports that you're paying for on a monthly basis. It's about time that you started making the most of those in the form of identifying what are the efficiencies within your ecosystem. First of all, what are the things, you know, which could be better utilized subsequently you have the opportunity to reduce the noise of a troubled tickets handle. It sounds pretty trivial, but as an average, you can imagine every shop is tickets has the cost in dollars, right? >>So, and there's so many tickets and there's desserts that get on a network and across an end-user application value chain, we're talking thousands, you know, across and end user application value chain could be million in a year. So, and so many of those are not really, you know, cause of concern because the problem is somewhere else. So I think that whole triage is an immediate cost saving and the bigger your network, the bigger the cost of whether you're a provider, whether you're, you know, the end customer at the end of the day, not having to deal with problems, which nobody can resolve, which are not meant to be dealt with. If so many of those situations, right, where service has just been adopted, which is coordinate quality, et cetera, et cetera. So many reasons. So those are the, those are some of the immediate cost savings. >>They are really, really significant. Secondly, I would say Raj mentioned something about, you know, the end user application value chain and an understanding of that, especially with this hybrid cloud environment, et cetera, et cetera, right? The time it takes to identify a problem in an end-user application value chain across the seven layers that I mentioned with the OSI reference model across network and security and the application environment, it's something that in its own self has a massive cost to business, right? They could be point of sale transactions that could be obstructed because of this. There could be, and I'm going to use a very interesting example. When we talk IOT, the integrity of the IOT machine is extremely pivotal in this new world that we're stepping into. You could be running commands, which are super efficient. He has, everything is being told to the machine really fast. >>We're sending everything there. What if it's hacked? And if that robotic arm starts to involve the things you don't want it to do. So there's so much of that. That becomes a part of this naturally. And I believe, yes, this is not just like from a cost saving standpoint, but anything going wrong with that code base, et cetera, et cetera. These are massive costs to the business in the form of the revenue. They have lost the perception in the market as a result, the fed, you know, all that stuff. So these are a couple of very immediate funds, but then you also have the whole player virtualized resources where you can automate the allocation, you know, the quantification of an orchestration of those virtualized resources, rather than a person having to, you know, see something and then say, Oh yeah, I need to increase capacity over here, because then it's going to have this particular application. You have systems doing this stuff to, you know, Roger's point your customer should not be identifying your problems before you, because this digital where it's all about perception. >>Absolutely. We definitely don't want the customers finding it before. So rich, let's wrap this particular question up with you from that analyst perspective, how can companies use make big impact quickly with AI? >>Yeah, I think, you know, and it has been really summed up some really great use cases there. I think with the, uh, you know, one of the biggest struggles we've always had in operations is isn't, you know, the mean time to resolve. We're pretty good at resolving the things. We just have to find the thing we have to resolve. That's always been the problem and using these advanced analytics and machine learning algorithms now across all machine and application data, our tendency as humans is to look at the console and say, what's flashing red. That must be what we have to fix, but it could be something that's yellow, somewhere else, six services away. And we have made things so complicated. And I think this is what it was. One was saying that we can't get there anymore on our own. We need help to get there in all of this stuff that the outline. >>So, so well builds up to a higher level thing of what is the customer experience about what is the customer journey? And we've struggled for years in the digital world and measuring that a day-to-day thing. We know an online retail. If you're having a bad experience at one retailer, you just want your thing. You're going to go to another retailer, brand loyalty. Isn't one of the light. It wasn't the brick and mortal world where you had a department store near you. So you were loyal to that cause it was in your neighborhood, um, online that doesn't exist anymore. So we need to be able to understand the customer from that first moment, they touch a digital service all the way from their, their journey through that digital service, the lowest layer, whether it be a database or the network, what have you, and then back to them again, and we not understand, is that a good experience? >>We gave them. How does that compare to last week's experience? What should we be doing to improve that next week? And I think companies are starting and then the pandemic, certainly you push this timeline. If you listen to the, the, the CEO of Microsoft, he's like, you know, 10 years of digital transformation written down. And the first several months of this, um, in banks and in financial institutions, I talked to insurance companies, aren't slowing. Now they're trying to speed up. In fact, what they've discovered is that there, obviously when we were on lockdown or what have you, the use of digital services spiked very high. What they've learned is they're never going to go back down. They're never going to return to pretend levels. So now they're stuck with this new reality. Well, how do we service those customers and how do we make sure we keep them loyal to our brand? >>Uh, so, you know, they're looking for modernization opportunities. A lot of that, that things have been exposed. And I think Raj touched upon this very early in the conversation is visibility gaps. Now that we're on the outside, looking in at the data center, we know we architect things in a very specific way. Uh, we better ways of making these correlations across the Sparrow technologies to understand where the problems lies. We can give better services to our customers. And I think that's really what we're going to see a lot of the, the innovation and the people really for these new ways of doing things starting, you know, w now, I mean, I think I've seen it in customers, but I think really the push through the end of this year to next year when, you know, economy and things like that, straighten out a little bit more. I think it really, people are going to take a hard look of where they are is, you know, AI ops the way forward for them. And I think they'll find it. The answer is yes, for sure. >>So we've, we've come to a consensus that, of what the parallels are of organizations, basically the cost of doing nothing. You guys have given some great advice on where some of those quick wins are. Let's talk about something Raj touched on earlier, is organizations, are they really ready for truly automated AI? Raj, I want to start with you readiness factor. What are your thoughts? >>Uh, you know, uh, I think so, you know, we place our, her lives on automated systems all the time, right? In our, in our day-to-day lives, in the, in the digital world. I think, uh, you know, our, uh, at least the customers that I talked to our customers are, uh, are, uh, you know, uh, have a sophisticated systems, like for example, advanced automation is a reality. If you look at social media, AI and ML and automation are used to automate away, uh, misinformation, right? If you look at financial institutions, AI and ML are used to automate away a fraud, right? So I want to ask our customers why can't we automate await oil in it, operation systems, right? And that's where our customers are. Then, you know, uh, I'm a glass half full, uh, clinical person, right? Uh, this pandemic has been harder on many of our customers, but I think what we have learned from our customers is they've Rose to the occasion. >>They've used digital as a key moons, right? At scale. That's what we see with, you know, when, when Huseman and his team talk about, uh, you know, network operational intelligence, right. That's what it means to us. So I think they are ready, the intersection of customer experience it and OT, operational technology is ripe for automation. Uh, and, uh, you know, I, I wanna, I wanna sort of give a shout out to three key personas in, in this mix. It's somewhat right. One is the SRE persona, you know, site, reliability engineer. The other is the information security persona. And the third one is the it operator automation engineer persona. These folks in organizations are building a system of intelligence that can respond rapidly to the needs of their digital business. We at Broadcom, we are in the business of helping them construct a system of intelligence that will create a human augmented solution for them. Right. So when I see, when I interact with large enterprise customers, I think they, they, you know, they, they want to achieve what I would call advanced automation and AI ML solutions. And that's squarely, very I ops is, you know, is going as an it, you know, when I talked to rich and what, everything that rich says, you know, that's where it's going. And that's what we want to help our customers to. >>So rich, talk to us about your perspective of organizations being ready for truly automated AI. >>I think, you know, the conversation has shifted a lot in the last, in, in pre pandemic. Uh, I'd say at the end of last year, we're, you know, two years ago, people I'd go to conferences and people come up and ask me like, this is all smoke and mirrors, right? These systems can't do this because it is such a leap forward for them, for where they are today. Right. We we've sort of, you know, in software and other systems, we iterate and we move forward slowly. So it's not a big shock. And this is for a lot of organizations that big, big leap forward in the way that they're running their operations teams today. Um, but now they've come around and say, you know what? We want to do this. We want all the automations. We want my staff not doing the low complexity, repetitive tasks over and over again. >>Um, you know, and we have a lot of those kinds of legacy systems. We're not going to rebuild. Um, but they need certain care and feeding. So why are we having operations? People do those tasks? Why aren't we automating those out? I think the other piece is, and I'll, I'll, I'll send this out to any of the operations teams that are thinking about going down this path is that you have to understand that the operations models that we're operating under in INO and have been for the last 25 years are super outdated and they're fundamentally broken for the digital age. We have to start thinking about different ways of doing things and how do we do that? Well, it's, it's people, organization, people are going to work together differently in an AI ops world, um, for the better, um, but you know, there's going to be the, the age of the 40 person bridge call thing. >>Troubleshooting is going away. It's going to be three, four, five focused engineers that need to be there for that particular incident. Um, a lot of process mailer process we have for now level one level, two engineering. What have you running of tickets, gathering of artifacts, uh, during an incident is going to be automated. That's a good thing. We shouldn't be doing those, those things by hand anymore. So I'd say that the, to people's like start thinking about what this means to your organization. Start thinking about the great things we can do by automating things away from people, having to do them over and over again. And what that means for them, getting them matched to what they want to be doing is high level engineering tasks. They want to be doing monitorization, working with new tools and technologies. Um, these are all good things that help the organization perform better as a whole >>Great advice and great kind of some of the thoughts that you shared rich for what the audience needs to be on the, for going on. I want to go over to you, give me your thoughts on what the audience should be on the lookout for, or put on your agendas in the next 12 months. >>So there's like a couple of ways to answer that question. One thing would be in the form of, you know, what are some of the things they have to be concerned about in terms of implementing this solution or harnessing its power. The other one could be, you know, what are the perhaps advantages they should look to see? So if I was to talk about the first one, let's say that, what are some of the things you have to watch out for like possible pitfalls that everybody has data, right? So yeah, that's one strategy, we'd say, okay, you've got the data, let's see what we can do with them. But then there's the exact opposite side, which has to be considered when you're doing that analysis that, Hey, what are the use cases that you're looking to drive, right? But then use cases you have to understand, are you taking a reactive use case approach? >>Are you taking quite active use cases, right? Or that that's a very, very important consideration. Then you have to be very cognizant of where does this data that you have vision, it reside, what are the systems and where does it need to go to in order for this AI function to happen and subsequently if there needs to be any, you know, backward communication with all of that data in a process better. So I think these are some of the very critical points because you can have an AI solution, which is sitting in a customer data center. It could be in a managed services provider data center, like, right, right. It could be in a cloud data center, like an AWS or something, or you could have hybrid scenarios, et cetera, all of that stuff. So you have to be very mindful of where you're going to get the data from is going to go to what are the use cases you're trying to, you have to do a bit of backward forward. >>Okay. We've got this data cases and I think it's the judgment. Nobody can come in and say, Hey, you've built this fantastic thing. It's like Terminator two. I think it's a journey where we built starting with the network. My personal focus always comes down to the network and with 5g so much, so much more right with 5g, you're talking low latency communication. That's like the two power of 5g, right? It's low latency, it's ultra high bandwidth, but what's the point of that low latency. If then subsequently the actions that need to be taken to prevent any problems in critical applications, IOT applications, remote surgeries, uh, test driving vehicles, et cetera, et cetera. What if that's where people are sitting and sipping their coffees and trying to take action that needs to be in low latency as well. Right? So these are, I think some of the fundamental things that you have to know your data, your use cases and location, where it needs to be exchanged, what are the parameters around that for extending that data? >>And I think from that point onward, it's all about realizing, you know, in terms of business outcomes, unless AI comes in as a digital labor, that shows you, I have, I have reduced your, this amount of, you know, time, and that's a result of big problems or identified problems for anything. Or I have saved you this much resource right in a month, in a year, or whatever, the timeline that people want to see it. So I think those are some of the initial starting points, and then it all starts coming together. But the key is it's not one system that can do everything. You have to have a way where, you know, you can share data once you've got all of that data into one system, maybe you can send it to another system and make more, take more advantage, right? That system might be an AI and IOT system, which is just looking at all of your streetlights and making sure that Hey, parent switched off just to be more carbon neutral and all that great stuff, et cetera, et cetera >>For the audience, you can take her Raj, take us time from here. What are some of the takeaways that you think the audience really needs to be laser focused on as we move forward into the next year? You know, one thing that, uh, I think a key takeaway is, um, uh, you know, as we embark on 2021, closing the gap between intent and outcome and outputs and outcome will become critical, is critical. Uh, you know, especially for, uh, uh, you know, uh, digital transformation at scale for organizations context in the, you know, for customer experience becomes even more critical as Swan Huseman was talking, uh, you know, being network network aware network availability is, is a necessary condition, but not sufficient condition anymore. Right? The what, what, what customers have to go towards is going from network availability to network agility with high security, uh, what we call app aware networks, right? >>How do you differentiate between a trade, a million dollar trade that's happening between, uh, you know, London and New York, uh, versus a YouTube video training that an employee is going through? Worse is a YouTube video that millions of customers are, are watching, right? Three different context, three different customer scenarios, right? That is going to be critical. And last but not least feedback loop, uh, you know, responsiveness is all about feedback loop. You cannot predict everything, but you can respond to things faster. I think these are sort of the three, uh, three things that, uh, that, uh, you know, customers are going to have to, uh, have to really think about. And that's also where I believe AI ops, by the way, AI ops and I I'm. Yeah. You know, one of the points that was smart, shout out to what he was saying was heterogeneity is key, right? There is no homogeneous tool in the world that can solve problems. So you want an open extensible system of intelligence that, that can harness data from disparate data sources provide that visualization, the actionable insight and the human augmented recommendation systems that are so needed for, uh, you know, it operators to be successful. I think that's where it's going. >>Amazing. You guys just provided so much content context recommendations for the audience. I think we accomplished our goal on this. I'll call it power panel of not only getting to a consensus of what, where AI ops needs to go in the future, but great recommendations for what businesses in any industry need to be on the lookout for rich Huisman Raj, thank you for joining me today. >>Pleasure. Thank you. Thank you. >>We want to thank you for watching. This was such a rich session. You probably want to watch it again. Thanks for your time.

>> Announcer: From London, England, it's theCUBE! Covering Coupa Insp!re'19 EMEA. Brought to you by Coupa. >> Hey, welcome to theCUBE. Lisa Martin on the ground in London at Coupa Insp!re'19. Can you hear all the buzz around me? You probably can hear it, it's electric. The keynote just ended, and I'm very pleased to welcome, fresh from the keynote stage, we have Rachel Botsman, author and trust expert from Oxford University. Rachel, welcome to theCUBE! >> Thank you for having me. >> Your talk this morning about the intersection of trust and technology, to say it's interesting is an understatement. You had some great examples where you showed some technology brands, that we all know, and have different relationships with: Uber, Facebook, and Amazon. And the way that you measured the audience is great, you know, clap the brand that you trust the most. And it was so interesting, because we expect these technology brands to, they should be preserving our information, but we've also seen recent history, some big examples, of that trust being broken. >> Rachel: Yeah, yeah. >> Talk to us about your perspectives. >> So what I thought was interesting, well kind of unexpected for me, was no one clapped for Facebook, not one person in the room. And this is really interesting to me, because the point that I was making is that trust is really, really contextual, right? So if I had said to people, do you trust on Facebook that you can find your friends from college, they probably would've clapped. But do I trust them with my data, no. And this distinction is so important, because if you lose trust in one area as a company or a brand, and it can take time, you lose that ability to interact with people. So our relationship and our trust relationship with brands is incredibly complicated. But I think, particular tech brands, what they're realizing is that, how badly things go wrong when they're in a trust crisis. >> Talk to me about trust as a currency. You gave some great examples this morning. Money is the currency for transactions, where trust is the currency of interactions. >> Yeah, well I was trying to frame things, not because they sound nice, but how do you create a lens where people can really understand, like what is the value of this thing, and what is the role that it plays? And I'm never going to say money's not important; money is very important. But people can understand money; people value money. And I think that's because it has a physical, you can touch it, and it has an agreed value, right? Trust I actually don't believe can be measured. Trust is, what is it? It's something there, there's a connection between people. So you know when you have trust because you can interact with people. You know when you have trust because you can place their faith in them, you can share things about yourself and also share things back. So it's kind of this idea that, think of it as a currency, think of it as something that you should really value that is incredibly fragile in any situation in any organization. >> How does a company like Coupa, or an Amazon or a Facebook, how do they leverage trust and turn it into a valuable asset? >> Yeah, I don't like the idea that you sort of unlock trust. I think companies that really get it right are companies that think day in and day out around behaviors and culture. If you get behaviors and culture right, like the way people behave, whether they have empathy, whether they have integrity, whether you feel like you can depend on them, trust naturally flows from that. But the other thing that often you find with brands is they think of trust as like this reservoir, right? So it's different from awareness and loyalty; it's not like this thing that, you can have this really full up battery which means then you can launch some crazy products and everyone will trust it. We've seen this with like, Mattel, the toy brand. They launched a smart system for children called Aristotle, and within six months they had to pull it because people didn't trust what it was recording and watching in people's bedrooms. We were talking about Facebook and the cryptocurrency Libra, their new smart assistants; I wouldn't trust that. Amazon have introduced smart locks; I don't know if you've seen these? >> Lisa: Yes. >> Where if you're not home, it's inconvenient for a very annoying package slip. So you put in an Amazon lock and the delivery person will walk into your home. I trust Amazon to deliver my parcels; I don't trust them to give access to my home. So what we do with the trust and how we tap into that, it really depends on the risk that we're asking people to take. >> That's a great point that you bring about Amazon, because you look at how they are infiltrating our lives in so many different ways. There's a lot of benefits to it, in terms of convenience. I trust Amazon, because I know when I order something it's going to arrive when they say it will. But when you said about trust being contextual and said do you trust that Amazon pays their taxes, I went wow, I hadn't thought of it in that way. Would I want to trust them to come into my home to drop off a package, no. >> Rachel: Yeah. >> But the, I don't know if I want to say infiltration, into our lives, it's happening whether we like it or not. >> Well I think Amazon is really interesting. First of all because so often as consumers, and I'm guilty, we let convenience trump trust. So we talk about trust, but, you know what, like, if I don't really trust that Uber driver but I really want to get somewhere, I'll get in the car, right? I don't really trust the ethics of Amazon as a company or like what they're doing in the world, but I like the convenience. I predict that Amazon is actually going to go through a major trust crisis. >> Lisa: Really? >> Yeah. The reason why is because their trust is largely, I talked about capability and character. Amazon's trust is really built around capability. The capability of their fulfillment centers, like how efficient they are. Character wobbles, right? Like, does Bezos have integrity? Do we really feel like they care about the bookshops they're eating up? Or they want us to spend money on the right things? And when you have a brand and the trust is purely built around capability and the character piece is missing, it's quite a precarious place to be. >> Lisa: I saw a tweet that you tweeted recently. >> Uh oh! (laughs) >> Lisa: On the difference between capability and character. >> Yes, yeah. >> Lisa: And it was fascinating because you mentioned some big examples, Boeing. >> Yes. >> The two big air disasters in the last year. Facebook, obviously, the security breach. WeWork, this overly aggressive business model. And you said these companies are placing the blame, I'm not sure if that's the right word-- >> No no, the blame, yeah. >> On product or service capabilities, and you say it really is character. Can you talk to our audience about the difference, and why character is so important. >> Yeah, it's so interesting. So you know, sometimes you post things. I actually post more on LinkedIn, and suddenly like, you hit a nerve, right? Because I don't know, it's something you're summarizing that many people are feeling. And so the point of that was like, if you look at Boeing, Theranos was another example, WeWork, hundreds of banks, when something goes wrong they say it was a flaw in the product, it was a flaw in the system, it's a capability problem. And I don't think that's the case. Because the root cause of capability problems come from character and culture. And so, capability is really about the competence and reliability of someone or a product or service. Character is how someone behaves. Character gets to their intentions and motives. Character gets to, did they know about it and not tell us. Even VW is another example. >> Lisa: Yes. >> So it's not the product that is the issue. And I think we as consumers and citizens and customers, where many companies get it wrong in a trust crisis is they talk about the product fix. We won't forgive them, or we won't start giving them our trust again until we really believe something's changed about their character. I'm not sure anything has changed with Facebook's culture and character, which is why they're struggling with every move that they take, even though their intentions might be good. That's not how people in the world are viewing them. >> Do you think, taking Boeing as an example, I fly a lot, I'm sure you do as well. >> Rachel: Yeah. >> When those accidents happened, I'm sure everybody, including myself, was checking, what plane is this? >> Rachel: Yeah. >> Because when you know, especially once data starts being revealed, that demonstrated pilots, test pilots, were clearly saying something isn't right here, why do you think a company like Boeing isn't coming out and addressing that head on from an integrity perspective? Do you think that could go a long way in helping their brand reputation? >> I never, I mean I do get it, I'm married to a lawyer so I understand, legal gets involved, governance gets involved, so it's like, let's not disclose that. They're so worried about the implications. But it's this belief they can keep things hidden. It's a continual pattern, right? And that they try to show empathy, but really it comes across as some weird kind of sympathy. They don't really show humility. And so, when the CEO sits there, I have to believe he feels the pain of the human consequence of what happened. But more importantly, I have to believe it will never happen again. And again, it's not necessarily, do I trust the products Boeing creates, it's do I trust the people? Do I trust the decisions that they're making? And so it's really interesting to watch companies, Samsung, right? You can recover from a product crisis, with the phones, and they kind of go away. But it's much harder to recover from what, Boeing is a perfect example, has become a cultural crisis. >> Right, right. Talk to us about the evolution of trust. You talked about these three waves. Tell our audience about that, and what the third wave is and why we're in it, benefits? And also things to be aware of. >> Yes! (laughs) I didn't really talk about this today, because it's all about inspiration. So just to give you a sense, the way I think about trust is three chapters of human history. So the first one is called local trust; all running around villages and communities. I knew you, I knew your sister, I knew whoever was in that village. And it was largely based on reputation. So, I borrowed money from someone I knew, I went to the baker. Now this type of trust, it was actually phenomenally effective, but we couldn't scale it. So when we wanted to trade globally, the Industrial Revolution, moving to cities, we invented what I call institutional trust. And that's everything from financial systems to insurance products, all these mechanisms that allow trust to flow on a different level. Now what's happening today, it's not those two things are going away and they're not important; they are. It's that what technology inherently does, particularly networks, marketplaces, and platforms, is it takes this trust that used to be very hierarchical and linear, we used to look up to the CEO, we used to look up to the expert, and it distributes it around networks and platforms. So you can see that at Coupa, right? And this is amazing because it can unlock value, it can create marketplaces. It can change the way we share, connect, collaborate. But I think what's happened is that, sort of the idealism around this and the empowerment is slightly tinged, in a healthy way, realizing a lot can go wrong. So distributed trust doesn't necessarily mean distributed responsibility. My biggest insight from observing many of these communities is that, we like the idea of empowerment, we like the idea of collaboration, and we like the idea of control, but when things go wrong, they need a center. Does that make sense? >> Lisa: Absolutely, yes. >> So, a lot of the mess that we're seeing in the world today is actually caused by distributed trust. So when I like, read a piece of information that isn't from a trusted source and I make a decision to vote for someone, just an example. And so we're trying to figure out, what is the role of the institution in this distributed world? And that's why I think things have got incredibly messy. >> It certainly has the potential for that, right? Looking at, one of the things that I also saw that you were talking about, I think it was one of your TED Talks, is reputation capital. And you said you believe that will be more powerful than credit history in the 21st century. How can people, like you and I, get, I want to say control, over our reputation, when we're doing so many transactions digitally-- >> Rachel: I know. >> And like I think you were saying in one of your talks, moving from one country to another and your credit history doesn't follow you. How can somebody really control their trust capital and creative positive power from it? >> They can't. >> They can't? Oh no! >> I don't want to disappoint you, but there's always something in a TED speech that you wish you could take out, like 10 years later, and be like, not that you got it wrong, but that there's a naivety, right? So it is working in some senses. So what is really hard is like, if I have a reputation on Airbnb, I have a reputation on Amazon, on either side of the marketplace, I feel like I own that, right? That's my value, and I should be able to aggregate that and use that to get a loan, or get a better insurance, because it's a predictor of how I behave in the future. So I don't believe credit scores are a good predictor of behavior. That is very hard to do, because the marketplaces, they believe they own the data, and they have no incentive to share the reputation. So believe me, like so many companies after, actually it was wonderful after that TED Talk, many tried to figure out how to aggregate reputation. Where I have seen it play out as an idea, and this is really very rewarding, is many entrepreneurs have taken the idea and gone to emerging markets, or situations where people have no credit history. So Tala is a really good example, which is a lending company. Insurance companies are starting to look at this. There's a company called Traity. Where they can't get a loan, they can't get a product, they can't even open a bank account because they have no traditional credit history. Everyone has a reputation somewhere, so they can tap into these networks and use that to have access to things that were previously inaccessible. So that's the application I'm more excited about versus having a trust score. >> A trust score that we would be able to then use for our own advantages, whether it's getting a job, getting a loan. >> Yeah, and then unfortunately what also happened was China, and God forbid that I in any way inspired this decision, decided they would have a national trust score. So they would take what you're buying online and what you were saying online, all these thousands of interactions, and that the government would create a trust score that would really impact your life: the schools that your children could go to, and there's a blacklist, and you know, if you jaywalk your face is projected and your score goes down. Like, this is like an episode of Black Mirror. >> It's terrifying. >> Yeah. >> There's a fine line there. Rachel, I wish we had more time, because we could keep going on and on and on. But I want to thank you-- >> A pleasure. >> For coming right from the keynote stage to our set; it was a pleasure to meet you. >> On that dark note. >> Yes! (laughing) For Rachel Botsman, I'm Lisa Martin. You're watching theCUBE from Coupa Insp!re London '19. Thanks for watching. (digital music)

>> Narrator: Live from the Wynn Hotel in Las Vegas, it's theCUBE covering Magento Imagine 2018. Brought to you by Magento. >> Hey, welcome back to theCUBE. We are continuing our coverage live from the Wynn Las Vegas at Magento Imagine 2018. We've had a really exciting day talking about commerce and how it's limitless and changing dramatically. Joining me next is Sucharita Kodali, the vice president and principal analyst at Forrester. Sucharita, it's great to have you on theCUBE. >> Thanks for having me, Lisa. >> So commerce is limitless. We've been hearing this thematically all day. You primarily are working with retailers on their digital strategies. And you've been doing this for a long time. Let's talk about the evolution that you've seen in the retail space with everybody expecting to have access to whatever they want to buy in their pockets. >> Right, right, right. I would say, so I've been working in the retail industry for the last two decades. I've been an analyst for the last 10 plus years. I've really seen a number of changes. And if I had to just summarize the biggest changes, one is just the inventory across different retail channels. So, that's definitely been a huge huge one. It's like, how do you, how do you order online, but then fulfill the item from a physical store or fulfill the item from another store? So those are, that's basically the digital transformation of retailers. Those are investments that companies like WalMart and Target have really been doubling down on and focusing on. The second big change is Amazon. And they single-handedly have transformed the retail industry. They have increased consumer expectations. And what Amazon's also done is reinvented retail as a business model. Because it is no longer about just selling product and being profitable selling that product. Amazon actually is not profitable with a lot of the items that it sells. It makes money in other ways. And it is probably what I would describe as America's first retail conglomerate. And that becomes a really interesting question for other companies to compete, do you have to become a retail conglomerate? Then, the third big change is just brand selling direct to consumer. I remember when I started at Forrester, my very first project was with a large consumer electronics company that asked, Well, should we even sell directly to consumers? There's channel conflict and issues with our distributors. And now, that's not even a factor. It's sort of table stakes you have to sell direct to consumer. And that's probably where we'll continue to see a lot of retail sales in the future. >> So the Amazon model, we expect to be able to get whatever we want whenever we want it, have it shipped to us either at home or shipped to us so we can go pick it up at a store. It's really set the bar. In fact, they just announced the other day that a hundred million Amazon Prime members. I know people that won't buy something if it's not available through Prime. But I think this morning the gentleman that was on main stage from Amazon said at least 50% of their sales are not products they sell, they're through all of the other retailers that are using Amazon as a channel as part of their omni-channel strategy. If you think of a retailer from 20 years ago, how do they leverage your services and expertise and advice to become omni-channel? Because as today, you said essentially it's table stakes for companies to have to sell to consumers. >> Yeah, yeah. There are so many questions that really require, I call it destroying the retail orthodoxies. And retail has historically been about buyers and merchandisers buying goods. There's the old expression in retail, You stack 'em high and watch 'em fly. And that is just where buyers would, Take a company like Toys R Us, they would basically take what Mattel and Hasbro told them to buy. They would buy a ton of it, put it in stores. And because there was less competition back in the '80s, consumers actually would buy that merchandise. And unfortunately, the change for retailers is that consumers have so much more choice now. There's so such more innovation. There are small entrepreneurs who are creating fabulous products, consumer tastes have changed. And this old paradigm of Mattel and Hasbro, or kind of fill in the blank with whatever vendors and suppliers, pushing things is no longer relevant. So, there was just an article in the journal today about how Hasbro sales were down by double digits because Toys R Us is now going to go out of business. So those are the kinds of things that retailers who did not adjust to those changes, they are the ones that really suffer. They don't find ways to develop new inventory, they don't find new channels for growth, and they don't protect their own. They don't build a moat around their customers like Amazon has done, or they don't find ways to source inventory creatively. That's where the problems are. >> You think that's more of a function of a legacy organization; having so much technology that they don't know how to integrate it all together? What do you think are some of the forcing functions old orthodoxies that companies that don't do it well are missing? >> Yeah, it's a lot of it is just in the old ways of doing business. So, a lot of it is being heavily dependent, for instance, on buyers and merchandisers buying things. I mean, one of the biggest innovations that Amazon realized was that, look you can sell things without actually owning the inventory. And that is, their entire, what we call the third party marketplace, and that is just so simple. But if you were to ask a buyer at a major retailer a decade or two ago, "Why do you have to buy the inventory?" their response would be, Well, you have to buy the inventory, that's just the way it is. And it's like, well why? Why don't you try to find a new way to do business? And they never did. But it took Amazon to figure that out. And the great irony of why so many retailers continue to struggle is that Amazon has exposed the playbook on how to sell inventory without owning it. And so few retailers to this day have adopted that approach. And that's the great irony I think, is that that's the most profitable part of Amazon's business is that third party marketplace. And every retailer I've talked to is like, Oh, it's really hard. We can't do that. But, the part of Amazon's business that everyone is looking to imitate is their fast shipping. Which, is the most expensive part of their business. Amazon is only able to afford the fast free shipping because of the third party marketplace. Other retailers want to get the fast free shipping without the marketplace. And it just doesn't make any sense. And that's really the heart of the challenge is that they just don't think about alternative business models. They don't want to change the way that they've historically run their businesses. And some of this could mean that merchants are not as powerful in organizations. And maybe that's part of the pushback is that, there could be a lot of people who lose jobs. The future will be robo-buyers and financial services you have robo-advisors, why not robo-planners in retail? >> So one of the keys then, of eliminating some of the old orthodoxies for merchants is to be able to pivot and be flexible. But it has to start from where in an organization from a digital strategy perspective? Where do you help an organization not fall into the Toys R Us bucket? >> Yeah, I think a lot of it does have to start with merchandising and putting in some interesting digital tools to help merchants be more flexible. So, you want to flex to supply and demand. And some of that comes with integrating marketplaces into your own experience. Some of it can be investing in 3D printers that can make things that are plastic or metals based on demand. That's something that I always wondered why Toy R Us didn't, for instance, make Fidget Spinners on demand. Why did you have to get them with a six month leave time from China, it never made any sense. You can scale service, so use technology to match great store associates with a customer who may have a question. And you don't have to be in the same store. It can be a Facetime call with somebody who is far away. But very few retailers do that. And finally, the last bit is really to look at new alternative business models and finding new ways of making money beyond just selling inventory. >> That's really key because there are so many oppurtunities when companies go omni-channel of not just increasing sales and revenue, but also reducing attrition, making the buying process simple and seamless. Everybody wants one click, right? >> Right. >> Super seamless, super fast, and relevant. It's got to be something if you're going to attract my business, you need to be able to offer something where you know me to a degree. >> Absolutely. >> Or know what it is I might have a propensity to buy. >> Absolutely. And that's the entire area of personalization. And that personalization can be anything from a recommendation that I give you. It can be proactively pushing a recommendation. That's what companies like Stitch Fix do is I tell you what I want and then they send you a box in the mail of things I think you would like and oh, by the way are your size and within your budget. It can be customization. One of Nike's most successful parts of their business is their Nike ID program which allows you to customize shoes according to colors and different sort of embellishments that you may like. And that's exactly the kind of thing that more retailers need to be looking at. >> What are some of the trends maybe that a B2B organization might be able to love or some of the conveniences that we have as consumers and we expect in terms of-- Magento, I was looking on their website the other day and a study that they've done suggests 93 percent of B2B buyers want to be able to purchase online. So, new business models, new revenue streams, but it really is a major shift of sales in marketing to be able to deliver this high velocity low touch model. What are some of the things that a business like a Magento, could learn from say a Nike with how they have built this successful omni-channel experience? >> Well, interestingly I think one of the most important things to recognize is that every B2B buyer is also a B2C buyer. And their expectations are set by their experiences in B2C. So, if you have everything from all of the information at your fingertips, all of that information is optimized for mobile devices. You have different ways to view that information, you have all of your loaded costs, like shipping, or tax, or if there's cross-border. All of the information related to the time to ship, any customs and duties, all of that needs to be visible because in any experience that you have with say a site like Amazon, you're going to get that information. So, the expectation is absolutely there to have it in any situation whether it's B2B or whether it's buying components or kind of very long tail items. That's basically the cost of doing business at this point, is that you have to deliver all of the information that the customer wants and needs. And if you don't, the customer is just going to opt to go purchase that product at whatever destination offers it. >> Somewhere else. >> And somebody will. That's the challenge when you have 800 thousand Plus eCommerce sellers out there selling every product imaginable in the both B2B and B2C landscape. >> So, on the data side there's so much data out there that companies have any type of business to be able to take advantage of that. I know that there's, BI has so much potential. Are you hearing retailers start to embrace advanced analytics techniques, AI machine learning, Where are they with starting to do that? I know that some eyeglass companies have virtual reality augmented reality type of apps where you can kind of try on a pair of frames. Where are you seeing advanced analytics start to be successful and help retailers to be able to target buyers that might say, oh, I can't try that on? No, I want to go somewhere that I can touch and feel it. >> Yeah, well, it's emerging still. I mean, retailers have a lot of data. I think they're trying to figure out where is it most useful. And one of the places where it is incredibly useful is in the backend with fraud management. So, after retailers were forced to put in chip cards as a payment form, what you started to see was more of the fraud shifting to eCommerce. I just had two credit cards that had to be shut off because of E-commerce fraud. But that is where you see the fraudsters going to. And what you see as a result of that is some innovators in that space technology companies really leveraging machine learning, AI, other advanced data techniques to identify fraudulent transactions and to better help retailers eliminate or reduce the percent of transactions that have to then be charged back. So, that's probably one of the most promising areas. There are others that are emerging. We're seeing more visual recognition technologies. House for instance, is excellent at that and Pinterest too. If there's part of an image you like you can click on it or you can tap it and see other images like that. And that's incredibly difficult. And it was even more difficult 10-15 years ago, but it's becoming easier. There's the voice element, voice to text or text to voice. I think that the best applications they're often in customer service, there are so many interactions that happen anywhere in a consumer facing world. It doesn't even have to be within retail. You can think about the complaints to the airline industry or to a bank. And a lot of it falls into a black hole. You always hear that oh, This call may be recorded, but it is really difficult to go back and transcribe that. And to really synthesize that into major themes. And what ML in particular can do is to basically pull out those themes, it can automate all of that, and can give insights as to what you could be doing, what you should be doing, what are the opportunities that you may not have even known existed. So there are definitely emerging places. I mean even a visual recognition, so we talked about House and Pinterest. Another great example is the computer vision that you have in the Amazon Go stores. And there's a robot that the Wal Mart stores are now testing to go find if there are gaps in the inventory that need to be filled. Or if something is running low or out of stock. So there are definitely some interesting applications, but it's still early days for sure. >> So last question, we've got to wrap here, but, we're in April 2018, what are some of the, your top three recommendations for merchants, as they prepare for say Black Friday coming up in what, six or eight months. What are you top three recommendations for merchants to be successful and be able to facilitate a seamless online offline experience? >> Well, we always have kind of imbalances between supply and demand, and that's where I do think things like third party sellers, third party marketplaces are huge. So to be able to leverage that is certainly one opportunity. Another is to think creatively about promotions. In Japan they have these promotions called Fukubukuro promotions, and it's basically like grab bags of like all the left over inventory. But then they basically put it into mystery bags where you can buy it for half off. And consumers line up around the block at stores to go buy these grab bags. Because they also have also like a gamified approach where, you know, one of out 10 of the bags will have like an Ipad or some really high value item. So people really like these things, and they have trading parties. So just new ways of having promotions beyond just the typical door busters that retailers think about. And then kind of third I think is just try to pace out the demand. One of the big issues in E-commerce has been just the burst in demand that always happen in December. And that creates a lot of problems from the standpoint of actually shipping the orders. So the more that you can pull those transaction forward into November, the better off you are from a fulfillment and supply chain standpoint. >> Alright Sucharita thank you so much for stopping by theCUBE >> Thanks Lisa >> And sharing your insights on the trends and what's going on in the commerce and E-commerce space. Really enjoy talking with you. >> Nice to talk to you too. >> We want to thank you for watching. You're watching theCUBE live from Magento Imagine 2018, I'm Lisa Martin. Stick around, I'll be back with my next guest after a short break. (upbeat music)