 OK, so good morning, everybody. I'm going to be talking today about something that as organizers of Big Data Spain, we keep seeing each year. We bring here some of the biggest experts in big data and artificial intelligence. They talk to you about the state of the art of artificial intelligence or big data. And we have many things, the feeling that you see this as something of the future, something that is the next big thing that should not be worrying you today. So I'm going to try to get you at least a little worried. And I'm going to start by talking about history and not the history of Spark or Hadoop or Big Data, real history. So in the 16th century, Spain ruled the world as it should be, right? So Philip II had the greatest army that the world had ever seen. And in fact, in 1588, he sent that fleet, which was called the invincible fleet. Nobody was supposed to beat it, to England, to invade England. If he had succeeded, the Spain would still be the ruler of the world. And this talk could be in Spanish, which would be much better for me. But he failed. In fact, he came up with the concept of epic fail. So it was one of the biggest catastrophes of military history. And seeing that, Philip II came up also with one of the greatest excuses in history, which is, I didn't send my ships to fight against the weather. So it was so good that many people remembered the excuse and not the fact that he invented the epic fail. So many companies are facing now a reality which is not so different from this, because the world is changing very, very fast. We see the kind of things that artificial intelligence are going to bring into every kind of business in the next years. And many organizations have not been built to resist these chains. So something must be done. And it's something that may cause the survival of the company itself. So there are only two options. Either you start looking for excuses right now, or you start making some changes. We are going to be talking about the second option. Of course, if you want a very good excuse, you already have the one from Philip II. I didn't send my company to fight against AI or this environment. But let's take a look about what is really being done outside today. What should worry me? What are others already doing with artificial intelligence? And there are a lot of fields that have real results today. One of the first ones can be personal life recommendations. I'm sure you are aware of this. This is pretty much the first field where artificial intelligence explodes. And it's something that really produces results. We are all aware that personalization works. 70% of customers prefer personalized shopping, cross-selling works. It can rise sales by 20%, profits by 30%. And we have also this canned laughter effect that I don't know if you know. Probably you all have watched a comic TV. And then you start hearing some laughter. And the minute you start hearing the laughter, you tend to laugh. You cannot stop yourself. You laugh even before knowing what you're laughing at. This is social anxiety. If you see the system is telling you that people who bought this have also bought this other thing, you will feel compelled to buy that other thing. So this is the things that are achieved with artificial intelligence in this arena. And this is something that has been done for years now. It's not something even from the present. It's something already in the past. I don't know if you are aware with this fact, some retailer many years ago, it's not even know the retailer, made an analysis, an automatic system that was going to cross-check every pattern of people buying products and come up with proposals of what products placed together in the shelves to rise up cross-selling. The system told a lot of things and one of them was put the beer besides diapers. And nobody understood why. They thought it was a myth from the data analytics system, but they did it and they checked that cells in both beers and diapers came up. And it was only later when they realized that what was happening was that first time parents went to the store to buy diapers and when they saw the diapers and the beer, they remember how two weeks ago they would be having beers with their friends and now they can't. So they would buy the six pack of beers to have the beer at home. So this is an example of something that is not intuitive or maybe even counterintuitive to humans, but machines are doing much better and boosting cells of those products. Same thing with this example of Target, also very well known and again, not of today. This was years ago. This result has already been achieved years ago. They did more or less the same thing. They set up an automatic system that was going to try to guess based on patterns of behavior of the users, what profile did they match? If they were first time parents, if they were divorces, or if they were people who had just bought a house and depending of this profiling, automatically the system would address them with specific offers of coupons. And then one day a parent called the company to Target to complain because his daughter was receiving coupons of first born produce. He said, okay, are you trying to get my daughter pregnant? She has, she's not even 18. So they investigated and at the end of the day, what happened was that the girl was pregnant and Target knew about that before her own family, which is exactly what can be done right now with artificial intelligence. So if we jump ahead to today, what are we seeing today in the field of recommendation and personalization? Probably one of the leaders is Netflix. Netflix is right now generating one billion revenue because of recommendations based on artificial intelligence. So they have 75 million subscribers and every one of them receive a personalized experience, a personalized recommendation that is right now the reason why 80% of TV shows are watched in Netflix. This is very, very important because they have also tested that in this area of TV channels, you have only 90 seconds to convince the user of something that's interesting for them. If not, they are going to abandon the TV. So in those 90 seconds, they have to recommend something interesting for the user. And it's a very, very complex scenario because it's very counterintuitive. In fact, they handle the term that against intuition, I'm going to recommend this show against this other one. How this works? They have a complete tagging of every minute, of every show that they have in their catalog and tagging, complex tagging that has been done by people who have watched every minute of the shows and they have a very profound profiling of every user of Netflix. When they connect, what shows they see, if they see them in a binge watch episode or they see week by week, and then they combine everything to make decisions, decisions that are not easy. What show should they recommend to a guy that connected yesterday? Something related of what he saw yesterday of what he saw one week before. So this is not something that humans can do. This is something that artificial intelligence is doing for Netflix and is doing this with this rate of success, one billion revenues for Netflix. Another example, we do stay in the field yet of personal life recommendation, but we moved to a food company. In this case, McDonald's, they launched an app APP to analyze customer behavior in some of the countries. This was, I think, in Netherlands, Japan, several others that covered 40 million end points. So this APP, mobile APP, received instant offers which were decided, depending on the behavior of the person and the weather where he was at. For example, if it's hot outside and I'm walking near a McDonald's store, I would receive an instant offer for a Sunday in that particular shop. So with this cross information and real-time offers, coupons, they got a result of 700% increasing redemption of these offers and a return to the stores twice and 47% of increasing expense. So you see that these are real results that are happening today have already happened. It's nothing of the future. More examples, Walmart, another retailer is storing every hour 2.5 petabytes which I'm a little lost of what is, but to give you an idea, it helped me to understand the magnitude of this number is more than everything that the humankind has written in his history. So you put together all the information in the books written in the history of humankind and it's still less than what Walmart is storing every hour. So they combine and analyze this information to anticipate customer needs, optimize operations and also even do facial recognition. They started with that as a kind of anti-theft mechanisms but now it's using it to measure, to gauge the frustration of their customers. So they can warn a customer representative about people getting frustrated because there are too many queues or whatever. So this again is something that is used today. But moving out of this field that may be more or less familiar with many of us, we will talk about real things that are being done today in other fields, not recommendations, personalization. For example, Netflix probably uses even more artificial intelligence to stream their contents than to personalize them because they have a huge problem and they need to serve content instantly, of course, to every one of the users and they need to do it from a lot of different content providers. So they have to select the correct one, the one that is going to work best for the user. And the behavior of the users is pretty bad known. I mean, you have there two examples of the throughput of a typical connection to Netflix. So Netflix needs to know what's going to be to your throughput in order to encode the content according to your capabilities but those capabilities look like that. So what to do, what they do is to analyze thousands and thousands of patterns of real throughputs correlated with devices, with locations, with times of the day to try to predict, to guess what's going to be your throughput. And based on that, they pre-encode the content that they are going to serve you. So they know better than you how your bandwidth is going to behave in the next second of minutes because they need to know that in order to encode the content accordingly. So again, a fantastic example of artificial intelligence in good use. More examples that we all know them but maybe not so well. We all know Amazon as one leader of artificial intelligence. We know about deep learning, about Amazon Go, that store where people don't have to check out because they are seen all the time. Amazon knows what they are doing and they are charged without more operations. We all know also about Amazon's recommendation algorithm. It drives, in fact, 30% of their total sales. And we probably know, I don't know if you know about this, the search by image functionality. In fact, it has been in the Amazon app for now some time but it's now integrated with the Snapchat and allows you to make photographs or videos or something that you are interested in and launch immediately to Amazon to buy it. These are pretty cool examples and really usable in fact, they are generating revenue but probably what is less known is the magic that is happening behind the scenes at Amazon Store. So we have there a completely automated process based on intelligent robots that are making a lot of decisions for Amazon from the time you click on the buy button. So from that point on, until you have the box at home, there is a complete race in order to serve you the package on the less time, the better. So what we have, we have an instant selection of the warehouse that is going to serve your order based on a lot of things. Of course, distance to your home but not only that, a stock, busyness of the clerks. If that warehouse is very busy, you may be redirected to other one that in that time has more clerks or they are less busy than the rest. They recognize defects automatically. So if the robot select a product and that product is defectuous, it will be returned to the store. You will not receive it at home and they are able with this system that we'll see in a minute to have 50% more inventory, 20% down in operating cost and 15 seconds to assemble one package. So at the end of the day, they click to ship, went down by 15 minutes, which is a lot. Take a look of how it works. Oh, we don't have sound. Well, this has captions at least. Well, I hope we get sound for the next one. But anyway, you have seen here how these robots, which are called Kiva robots. This was an acquisition from Amazon some years ago. Make all the decisions for the pickers. In fact, what is called the picker, which is the Amazon guy who stands there in the storage years ago, used to actually pick something, to move to the shelf and pick something. Now the robots are bringing stuff to them. So they pick the shortest route to the nearest clerk and they bring everything to them. And when the clerk receives the product, he also is receiving the correct box and the correct measure of tape to box it. So actually, the time that they spend handling the package is very, very small. So everything else is automated and it's not a matter of automation itself. The matter here is the intelligence that drives that automation so that the decisions that they take, these robots are the correct ones, are the more efficient ones in order to maximize the results. Do we see there how they are pushed into the correct tracks, the ones that are going to be closer to your home in order to complete the order in the less time possible. This is what all retailers now are competing with and that's why it's so hard to compete with this. By the time you take an Amazon delivery, by the time you have opened it, you have spent more time in contact with that package than the whole personnel of Amazon. So it's that efficient, okay? More examples, augmented emotional intelligence. We have heard a lot and we will keep hearing a lot about chatbots, which probably will be the next big thing in artificial intelligence, the possibility of really substituting customer care with expert systems, with automatic systems that talk to you and solve the problems to you. But the jury is still out if that's a reality or that's something for, let's say, next year. But I wanted to show you something that is a reality one now, which is what is called augmented emotional intelligence, which is the ability to help real people to make better decisions when they are serving the client. Let's see if this video can be heard, I guess not. Well, I will try to describe it for you when we see it. But the thing here is that when the customer representative is speaking with a client, in real time, there is an analysis of what is being said and an emotional context that is going to be displayed in order to get those advices that we are seeing on the radar, that you are speaking quickly, you are talking a lot, you are not generating enough empathy with your client, try another approach which you see there, empathy cues, and on the same time, we are analyzing everything that is being said. So these artificial intelligence systems are accompanying you when you are serving the client in order to give you real time tips about how to improve the experience of that particular client. So we are going to combine an analysis of what you are saying with an analysis of how you are saying it, and we are going to try to correct you, to improve your behavior for the next interaction with the client, and also to shift that call to a supervisor if we see that it's not being handled properly. So we have that option. Okay, let's see if we can recover the sound later. So more examples, shifting to different areas of applications that are really being used today. Fraud control, this is also a very big area for artificial intelligence and something that is really producing results, and very interesting results. One very good example would be PayPal which is processing, as you may know, $500 billion a year in payments, so for us, for them controlling fraud is a very big thing. What they do is each time that they detect fraudulent operation, they make a pattern of it, and they apply it, they train a model in order to avoid that globally that specific fraud is repeated anytime in the future. So right now they have thousands of those patterns that are compared in real time to each operation that you do. And I mean in real time, even before it happens, that's why I compare it to minority reports. You probably have seen that moving of the pre-crime thing that avoided crime even before it happened. So this is kind of the same. Artificial intelligence allows these kinds of banks and payment organizations to detect an anomaly, detect a pattern of something that is going to be fraudulent before it happens. So we can avoid it and not wait until it's too late. Right now in PayPal, they have spectacular results. They have gone down from 1.32% of fraud against revenue, which is the average in the sector, to 0.32%. So you see the advantage that they got from using this kind of artificial intelligence systems. In fact, I don't know if you know that they have acquired just some months ago a company, a specific company, for this artificial intelligence fighting about fraud, which is called Simility, for 120 million. So this is a niche, this is an area in which artificial intelligence is used, and is used in a very interesting way, in a very productive way. More examples, shifting from the personalization and recommendation wall, which is something that may be familiar to you, to more obscure areas in which probably you are not aware of how much is being done already. We have the energy and utilities areas. Here we are talking about forecasting energy for the big distributors, the capacity of predicting energy demand so that they can control the supply and optimize efficiency in order to serve the demands when something happens globally, in a city, in a center, whatever. To energy efficiency. Here the best example is Google DeepMind. Google DeepMind is a company that Google acquired a couple of years ago or more, that they specifically wanted to boost efficiency of the data centers, the cost of the energy used in their data centers. So this is an artificial intelligence system that analyzes everything in the machines, in the behavior of every component of the data center and starts patterns to boost efficiency. So they say, Google says that it reduced cost by 40% once the machine learning system was finished. You can see there an example of how when these data models once trained were put in production, the savings that they generated immediately compared with the previous energy waste. And then we have the energy accessibility front, which is kind of translating this, shifting this to the user, to the end user. So how can every one of us profit from the same level of artificial intelligence? So we can take a look at how these guys do, and we have to give you the pitch because we can hear the sound. But basically you put a device, just as you see, grabbing your electrical input and from the moment you plug in, it's going to sample your electrical input by millions of time a second, which is of course much more than whatever system not based in artificial intelligence can do. So with this information, this artificial intelligence system is going of course to give you a real time situation of how much you are spending but not only that, as you turn on and off the different device of your house, it's going to detect a pattern, the pattern of behavior of these devices and identify them so that you can recognize which one of your devices is the one making the waste and receive usually insights about when to plug on or off the different devices. And of course it also recognizes when one of them is not behaving properly. If the test failure before you even realize you have some problem there. So again, something has been done today. And manufacturing going farther from the user to the back and to the biggest companies, so to speak. We can talk about manufacturing in two of the biggest companies there, Siemens and General Analytics, which are doing pretty much the same. We see here the Siemens example, they launch MindSphere as a global system that monitor records and analyze everything in the supply chain from manufacturing, design delivery to find problems and to provide solutions even before humans know that there are problems involved. This means a huge insight, a huge assimilation of every data that is being generated and acting on it. This is so huge that the German government came up with the term industry 4.0 that probably you have heard of because of this, because of these advances. So we can again hear it, but let's take a look at what they do is basically to digitalize everything that is happening in every manufacturing site that they have so that they receive instantly data from all this particular system and gain insights about them. So the idea here, the holy grail of these systems is maintenance, is the possibility of finding out about a piece that is going to be grown in a matter of days or minutes because there is a huge spend of money in these kind of companies in regular maintenance. Having a guy that periodically goes once a year or once a month to an installation to check if everything is right. So if you can save that cost and instead get warning beforehand that this happens, you will stop downtime practically to zero and you will save thousands and thousands of years in this personal that has to go to those sites. Same thing or more or less the same thing is trying to do or is doing general electric with this, what they call the billion manufacturing suite that they deploy in their factories in order to build a whole intelligent system that links everything, design, engineering, delivery, distribution, everything is correlated, everything constitute a sentient being that is going to monitor everything that's happening. They deployed in the first factory that they deployed, it was in India, they improve efficiency 18% and they are processing now one terabit every day of this kind of data and doing this kind of the analysis. So again, something that is being used today. And to finalize, I wanted to show you another example to maybe take it out of the common path. This is something, it's a pity that we can hear it, but it's an example of artificial intelligence applied to agriculture. So this is an expert system that is going to recognize weeds or elements in a plantation that needs to be removed or treated in a special way. So what they did was have cameras in this kind of, I don't know the name of the truck, but they are going to be exploring everything that is on the ground and train it, train a model with what a weed looks like, what the different types of weeds look, which kind of safety area they need to live around each of these plants in order to put herbicide on them, et cetera, and once the model is trained, what they do is to launch these trucks on their own so they can apply herbicide more or less or apply good products to the good herbs and they are saying here that they reduce by 10 the personnel that they need to take care of these cotton plantations, which is what we are seeing here. So again, it's something very creative and that's why I'm showing it to you. It's a creative use of artificial intelligence in an industry or in an area that probably you would have thought that has nothing to do with technology, with artificial intelligence or with this big data. We train models of recognizing different types of herbs and from that point on, we are going to save products, to save personnel, and to boost productivity enormously. So since we can't hear that, I will not dig deeper in those videos, but we will talk about what to do to be one of these. So I hope I can have you have convinced that there are real big, big results of artificial intelligence being used today, not tomorrow, and even with that, most of you will live here tomorrow, probably with one of two syndromes that are very common. So let's see if you recognize this yourself in any of these syndromes. First of all, this is very good, Amazon does it very good, Google, Netflix, but I'm not one of them. I'm in fact fed of hearing about Amazon, about Netflix, about artificial intelligence. So next time somebody tells me I have to apply artificial intelligence, I will shoot him in the face. I'm fed of this. It does not work for me. So this is the first syndrome. And it's really something that you should be happy about because if you saw here today and tomorrow use cases that apply to your company, then it would be already too late for you because at the end, when this goes to the maturity, either you are different or you are cheaper. If you do not see here nothing that really applies to you, then you have the time, you have the space to come up with something that is different from your competitors because every innovation goes through these cycles, through this curve. We are probably halfway through this circle. So at the beginning it's very difficult to stand out but you have less competition in the market. Halfway through it is still very difficult to stand out or more difficult, this is the effect of nothing that I see really solves my particular problem but you start having more competitors that are solving those problems. And by the time that you come here to be at Spain and say, okay, that's the solution to my problem, you will be in the saturation phase when the thing you see will be in everybody's system. So you cannot differentiate with that. It will only be a matter of being cheaper than the other ones, which is something that we don't want. If you see the image behind the graphic, it's a commercial from BMW, which they realized this a lot of years ago and they shifted tracks in order to not sell their cars because their cars were at the end of the day the same product than the other manufacturers and shift to an emotional message, which is something where they could differentiate with the competition. So this is an example of shifting tracks in order to be different, okay? Second syndrome, if it works, don't change it, right? So why should I change things that are working perfectly in my company and have been working for years? So this is the Blackberry syndrome, blockbuster, AOL, all those guys thought that are at a particular time. But I'm guessing that you are still thinking, okay, but this is the plain crash effect. That's not going to happen to me. This only happened to these guys because they were unlucky. But see if you recognize yourself in this story that this is something that is shown in many MBAs. The problem of not questioning your plan. At the time, the Space Shuttle, the American Space Shuttle was the most advanced vehicle in the world. But it lacked because of the effects in the design that were due to the original design of Roman chariots, Roman word chariots 2000 years before. Why? Because those propellers that you see needed to be transported from the factory which was in Utah to Cape Canaveral to the lounge site. And they had to be transported through these tunnels that have had a particular width. And you may think, okay, somebody will have designed those tunnels with the intelligence to know what's the proper width. But no, the width came from the width of the rail tracks which was a very weird number, four feet and eight and a half inches. Why? Because they simply used the plans of the English immigrants that came to America hundreds of years before. And these guys simply reuse the plans of the tram tracks that were used years and years before. The tram guys reuse the plan of the tram wheels, the wheel axis that were again reused from the plans of the horse wagons which went hundreds of years before. Why the horse wagons had this special width? Because they had to go through English roads that had routes, had routes made by the Roman word chariots hundreds of years before which are the only ones that in 2000 years put intelligence in that width, four feet and eight and a half inches. It's the width of two Roman horses side by side. So you can actually say that the space shuttle was designed like this, like this of two war Roman horses specifically. And still it works. So you may think why it's grown, it works. It was in fact the most advanced vehicle in the world but it could be better. That's the thing, we are in a space when we do not think it's to work. We need them to be the best in order to compete because the market allows us to be the best and we have to fight for that. Okay, so at the end of the day, you need to innovate. You need to put everything that is circulating here, these new things that everybody else is doing to put, you need to put them to good use in your business, in your organization. And before this, you need an innovation framework. Something that is not okay, I'm going in the next break before the next talk, I'm going to think how I can use artificial intelligence in my business. No, you need to have a mechanism in track in your company to create, test, scale, and renew your ideas. You need to have a thesis about innovation. You need to start from your particular vision about how your business is going to go in the next years and what's my innovation strategy for the next years because I need to have a goal, I need to have a horizon in order to go there. And then I have to deploy an innovation funnel, a mechanism running all the time that gets me from idea generation, which is the hardest part of all, to testing those ideas, scale them, and finally, if they work, adding them to the core. And you will notice that I have not talked about technology. Innovation, you could say that has nothing to do, in fact, with technology. You have to have ideas first, and then comes technology. So you have also to keep track of how good you are doing with your plan. I mean, to have these families good, but you need to measure every quarter, every year, if you are producing results, if really your investment match these adjacent of transformational products that are going to reinvent your business because innovation is not incremental, it's revolutionary. You do, it's not enough to do small steps from where you are. You see the kind of changes that these guys that I showed you are doing in their business. Do you need to come up with something that is comparable to those kind of changes? Okay. If you are questioning how far am I from this scenario, you can take this little quiz. You don't have to do it in public and raise your hands. Don't be embarrassed, but you ask yourself these questions. How is your focus? You focus on cost reduction, or you focus on operational efficiency, or you focus on exploring new opportunities and markets. What's your budget, annual, departmental, or is based on products, like the venture capitalists do? Your processes are linear, are pre-planned, or are iterative, experimental? And the hardest of all the question, in your company, what's the culture about failure? It's prohibited, is something bad, or is something that is assumed as necessary? So these are the questions that you should be making yourself to see how are you close enough to the innovation. Okay. So how to get there? If you fail the test, tips that can help you to pass it next time, from our experience, helping clients to get there. First of all, add digital agitators to the mix. You cannot pretend to know about everything. You have your business, you have your experts in your business, and there are people who are focused exactly in this, in having the experience, the skills to have these ideas and to really revolutionize, reinvent your business. So have some of them in your team. First, second, start thinking from inside, from outside instead of inside out. What I mean with this, many companies have grown so complex, so tangled inside, that they start translating their complexity to the user, to the end user. For example, there are retailers that have food and non-food organizations, because internally, the logics are different, the personal is different, but for the end user who is making an order, there is really no difference. So you are really translating with this complexity something that is only your problem to the user. So you should start thinking the other way around what the user needs for me and how should I be organized to comply with those needs. Use co-creation dynamics. The reality is that most of you, all of you have the knowledge to be different, to be unique. That knowledge is only in you, but it's usually hidden. You need to encourage the creativity of the people that needs to come up with this revolution idea. And this is something that is already invented. I mean, this co-creation dynamics makes you go through these diverse, diverse cycles and they allow you to come up with ideas that are at the back of your heads, that are coming together with this co-creation dynamics, are going to flourish and to get into this innovation funnel that is going to end up in real industrial application of those ideas. So do this. Get your team together and this is something that may be a challenge. Probably in a big company, in a company that has been going on for decades, your super heroes, your IT, your management, your business people are there, but are straight to different locations. They don't talk to themselves anymore and it's difficult to get them together. I compare this to the scenario of the Blues Brothers. I don't know how old am I and how many of you know about the Blues Brothers movie, but this was a movie where these two guys belong to an old band, an old Blues band, and they were going after every member of the band, saying we are on a mission from God, we are putting the band together. So that's something that you must do in order to get your heroes back together because normally they will be far away in your company and they need to be close and working together. So get your team together and be very, very fast. So here the quote that summarized all this came from Seth Godin that said, fail fast, fail cheap, fail often, and fail in a way that does not kill you. So you need to promote failure and it's something that may seem counter-intuitive in many organizations, but it's the only way to come up with the velocity that these other guys have. Amazon, Google, every example that we saw before have engineers testing idea thousands of times a day, thousands of tests to see which one works. So you need to be very, very fast to move on part with them and have your own ideas working. If you are not embarrassed by your first version of your product, by your first release, you've launched too late. This is what the LinkedIn co-founder said and it's something that we push again and again to our clients. You need to be fast, you need to get out there and see if your idea, if your product is doing good before making more decisions. You need to shorten your cycles. And finally, and that's my last tip, be brave. I mean, this is a tremendous era for all that we are here. We are privileged. We are being witnesses of the first IT revolution in 30 years. We have not seen anything like that, like what we are seeing now, like artificial intelligence, since probably the internet. So we are in a position to do something, but it must be something revolutionary, not incremental. You have to think how to grow wheels and get out there and do things that are really revolutionary. Coming back to the beginning, I would say that you have to auction and choose the right one. Don't get out here today saying, okay, that's all very good, but in my case, I have a lot of problems. This can't be done. My boss is not going to allow it. The technology is not ready. Do you know what? Let's wait for one year. That's making excuses. So stop making excuses and start making history in your company. You have the chance. So take it. Thank you. It's the time for the question. Okay. Any question? I have one. Yeah, why didn't they all your sound? I have one. If no one have one, I can ask you. Okay. Would you prefer? Let them or I start. No, no, let them see if there is something. I can see it. So it's complicated. Go ahead. You say that innovation is not about technology. I like that part, but not have a whole track about technology here at BDS. Isn't technology a very big part of this? Of course it's a very big part of this. And that's why Big Data Spain has a technology track. But I would say again that that's not the problem of most companies in order to be competitive, in order to come up with these ideas. Technology is there for us. We don't have to launch into these technology proof of concepts without thinking why we are doing it. Because it's there for us. The problem that we have is a business problem. How I am going to use that technology in order to be different, in order to have a product that is ten times better than what my competitors have. So once I have that question answered, then I can go to technology. But I can assure you that you are not going to have a problem in the technology side. That's why it's true that I did not mention technology almost in any time because it's not the important key here. We're going to be out of time. So there is no time for more questions. If you have more questions, you can ask him. Right now we have ten minute break. You go outside, you take something, you drink, you get enjoyed, and in ten minutes later, you come back. Because our next guest, she came from the stock home. She's going to talk about the privacy, the policy. I'm in his company in Spotify and he's a really sweet guy, girl in this case. So I hope you soon in ten minutes. So be back. Thank you. Thanks to you. Thanks.