>> Announcer: Live from Munich, Germany, it's the CUBE, covering DataWorks Summit Europe 2017. Brought to you by Hortonworks. >> Hey welcome back everyone. We're here live in Munich Germany for DataWorks 2017 Summit, formerly know as Hadoop Summit, now called DataWorks. I'm John Furrier with the CUBE, my co-host Dave Vellante, here for two days of wall-to-wall coverage. Our next guest is Nadeem Gulzar, head of advanced Analytics at Danske Bank. Welcome to the CUBE. >> Thank you. >> You're a customer but also talking here at the event, bringing all your folks here. Your observation, I mean, Hadoop is not going away, certainly we see that. But now, as John Kreisa, who was MC'ing, was on earlier said, open up the aperture to analytics, is really where the action is. >> Nadeem: Absolutely. >> Your reaction to that. >> I completely agree, because again, Hadoop is basically just the basic infrastructure, right. Components build on components, and things like that. But, when you really utilize it, is when you add the advanced analytics frameworks. There are many out there. I'm not going to favor one over another. But the main thing is, you need that to really leverage Hadoop. And, at the same time, I think it's very important to realize how much power there actually is in this. For us at, in Danske Bank, getting Hadoop, getting the advanced analytics framework, has really proven quite a lot. It allowed us actually to dig into our core data, transaction data for instance, which we haven't been able to for decades. >> So take me through, because you guys are an interesting use case because you're advanced. You're gettin' at the data, which is cutting edge. But you're going through this transformation, and you have to because you're on the front lines. Take us inside the company, without giving away any trade secrets, and describe the environment. What's the current situation, and how is it evolving from an IT standpoint, and also from the relationship with the stakekholders in the business side. >> So again, we are a bank with 20,000 employees, so of course in a large organization you have silos, People feeling okay, this is my domain, this is my kingdom, don't touch it. Don't approach me, or you can approach me, talk to me, you have to convince me, otherwise don't talk to me at all. So we get that quite a lot, and to be honest, from my point of view, if we do not lift as a bank, we're not going to succeed. If I have success, if my organization of almost 60 people have success, that's good in itself, but we are not going to succeed as a bank. So for me, it's quite important that I go down and break down these barriers, and allow us to come in, tell the business units, tell them what sort of capabilities do we bring, and include them. That is actually the main key. I don't want to replace them or anything like that. >> So an organizational challenge is to get the mindset shifted. How 'about process gaps and product gaps? 'Cause I mean I almost see the sequence, kind of a group hug if you will, organizational mindset, kind of a reset or calibration. And then identify processes and then product gaps, seem to be the next transition. >> Absolutely, absolutely, and there are some gaps. Still, even though we have been on this journey for a considerable amount of time, there are still gaps, both in terms of processes and products. Because again, even though we have top management buy in, it doesn't go through all the way down to the middle layer. So we still struggle with this from time to time. >> How do you break down those barriers? What do you do, what's your strategy? >> I'm humble, to be honest. I go in, I tell them, listen you guys I have some capabilities that I can add to your capabilities. I want you to leverage me to make your life easier. I want to lift you as an organization. I don't care about myself, I want you to be better at what you're doing. >> So Nadeem, the money business and the technology business have always had a close relationship. It was like in 2010 after we came out of the downturn, it was like this other massive collision. You had begun experimenting with Cloud, the shift, CapEx to OpEx. The data thing hit in a big way, obviously mobile became real. So talk about the confluence of those technologies, specifically in the context of your big data journey. Where did you get started, and how did it evolve? >> So actually it fit in quite nicely because we were coming out of this down period, right, so there was extreme amount of focus on cost. So, of course at the time where we wanted to go into this journey, a lot of people were asking, okay how much does this cost, what's the big strategy, and so on. And how's the road map going to look like, and what's the cost of the road map? The thing is, if you buy some off the shelf commercial product, it's quite expensive. We can easily talk like half a billion, something like that, for a full end to end system. So with this, you were allowed, or we were allowed, to start up with relatively small funding, and I'm actually talking about just like a million dollars, roughly. And that actually allowed us a substantial boost in the capability department, in allowing us to show what kind of use cases we could build, and what kind of value we could bring to Danske Bank. >> So you started with understanding Hadoop? Is that right, was that the starting point? >> Yes, in a fairly small, very researched team set up. We did the initial research, we looked at, okay what could this bring? We did some initial, what we call, proof of value. So small, small, pilot projects, looking at, okay this is the data. We can leverage it in this way, this is the value we can bring. How much can we actually boost the business? So everything is directly linked to business value. So, for instance, one of the use cases was within customers, understanding customer behavior, directly linking it to marketing, do more targeted marketing, and at the end get more results in terms of increased sales. >> We just started a journey 2009, 2010, is that right? Or was it later? >> No, we started somewhat later. The initial research was in '14. >> In '14? Okay, alright, so '14 you sort of became familiar with Hadoop, and then I imagine, like many customers, you said okay, wow this stuff is complicated, but you were takin' it in small chunks, low risk. Let's get some value. Marketing is an obvious use case. I would imagine fraud is another obvious use case. So then, how did that evolve? I mean it's only a few years now, but I imagine you've evolved very quickly. >> Extremely quickly. Actually, within two months of the research, we actually saw a huge benefit in this area, and directly we went with the material to the senior members of the different boards we wanted to affect, and actually, you could call it luck. But, maybe we were just well prepared and convincing, so we actually directly got funding at that point in time. They said, listen, this is very promising. Here you go, start off with the initial, slightly larger projects, prove some value, and then come back to us. Initially they wanted us to do two things, look into the customer journey, or doing deeper customer behavior analytics, and the second was within risk. Doing things like, text mining, financial statements, getting some deeper into that, doing some web crawling on financial data such as Bloomberg, etcetera, and then pull it into the system. >> To inform your investments as a financial institution. From an architecture and infrastructure standpoint, we talked about starting at Hadoop. Has it evolved, how has it evolved? Where do you see it going? >> It has evolved quite a lot in the past couple of years. And again, to be honest, it's like every quarter something new is happening and we need to do some adjustments even to the core architecture. And with the introduction of HDB 3 hence later this year, I think we're going to see a massive change once again. Hortonworks already calls it a major change, or a major release. But actually, the things they are doing is extremely promising, so we want to take that step with them. But again, it's going to affect us. >> What's exciting about that to you? >> The thing that's very exciting is, we are now at like a balance point, where we have played quite a lot, we have released a couple of production grade solutions, but we have really not reached the full enterprise potential. So getting like into the real deep stuff with living under heavy SLA's, regulation stuff. All these kind of things is not in place yet, from my point of view. >> We talk a lot about, in the CUBE, and in our company, about these emergent work loads; you had batch, interactive, and the world went back to batch with Hadoop, and now you have this continuous workload, this streaming real-time workloads. How is that affecting your organization, generally, and specifically, you're thinking about architecture. How real is that and where do you see that in the future? >> It's the core, to be honest. Again, one of the main things we are trying to do is look into, so, gone are the days with heavy, heavy batches of data coming in. Because if you look at Weblocks for instance, so when customers interacts with our web, or our tablet solution, or mobile solution, the amount of data generated is humongous. So, no way on earth you can think about batches anymore. So it's more about streaming the data all the way in, doing real time analytics and then produce results. >> What would you say are your biggest, big data challenges, problems that you really want to attack and solve. >> So, what I really want to attack is, getting all sorts of data into the system. So, you can imagine, as a bank we have 2,000 plus systems. We have approximately 4,000 different points that delivers data. So getting all that mass into our data link, it's a huge task. We actually underestimated it. But now, we have seen we have to attack it and get it in because that is the gold. Data is the future gold. So we need to mine it in, we need to do analytics on top of it and produce value. >> And then once you get it in there, I'm sure you're anticipating that you want to make sure this doesn't go stale, doesn't become a swamp, doesn't get frozen. It's your job to talk about data oceans, which is really the long term vision I presume, right? >> And that is a key as well because with the GDPR for instance, we need to have full mapping and full control of all the data coming in. We need to be able to generate metadata, we need to have full data lineage. We need to know what, all the data where it came from, how it's interconnected, relations, all that. >> And that's what, two years away from implementation? Is that about right? >> It's going to take a while, of course. But again, the key thing is we make the framework so all the data coming in step by step, has that. >> Yeah, but so GDPR though, it goes into effect in '19, is that correct? >> It's actually May '18. >> May '18, oh, so it's much tighter time frame then I realized. >> John: You're under the gun. >> Nadeem: Yes. >> Okay, observation here at this event, obviously a lot of IOT, for you that's people. People and things are kind of the edge of the network. The intelligent edge is a big, big topic. Very dynamic. >> Nadeem: Extremely dynamic. >> A lot of things happening. Lot of opportunities for you to be this humble service provider to your constituents, but also your customers. How do you guys view that? What's the current landscape look like as you look outside the company and look at what's happening around you, the world. >> A lot of cool things are going on, to be honest. Especially in IOT, right? I mean, even though we are a core bank, still, there are a lot of sensors we can use. I talked a bit about, under the keynote, about ATM's, right? So, we're also looking at how can we utilize this technology? How can we enable our customers? If you look at our apps, they also generate extreme amounts of data, right? The mobile solution that we have, it gives away GPS location and things like that. And we want to include all that data in. At the end of the day, it's not for our gain, we are not always looking at making the next buck, right? It's also about being there for the customer, providing the services they need, making their banking life easier. >> And your ecosystem is evolving and rapidly adding new constituents to your network because, then you have the consumer with the phone, the mobile app alone, never mind the point of sale opportunity at the ATM. Now a digital, augmented reality experience could be enabled where you now have fintech suppliers, and potentially other suppliers in this now digital network that could be relational with you. >> Yes, and our job is to make sure that we leverage that. Acquiring a banking license is extremely difficult. But we have it, and what we need to do is to engage these fintechs, partners, even other banks, and say listen guys we invite you in. Utilize our services, utilize our framework, utilize our foundation and let's build something upon that. >> If you had to explain, Nadeem, this fintech start up trend because it is super hot, what is it? I mean how would you describe to someone who's not in the banking world. 'Cause most people would be scratching their head and say, isn't that banking? But, now this ecosystem is developing of new entrepreneurial activity and they're skyrocketing with success 'cause they have either a specialty focus, they do something extremely well. It may or may not be in a direct big space with a bank, but a white space. Use cases. So, is it good? Is it bad? Is it hype? What's the current state of the fintech situation? >> From my point of view, it's awesome. And the reason is, these guys are pushing us. Remember, we are a hundred fifty plus year old bank. And sometimes we do tend to just pat on our back and say, okay, this is going good, right? But, these guys are coming in, giving some competition, and we love it. >> Give me an example of a fintech capabilities. Randomly bring up some examples to highlight what fintech is. >> So what we've seen in, for instance the German market, is the fintechs coming in, utilizing some of the customer data, and then producing awesome new applications. Whether it is a new net bank, where a customer can interact with it, in a much, much more smoother way. Some of the banks tend to over clutter things, not make it simple. So things like, where you can put in, you can look at your transactions in a Google Map, for instance. You can see how much do you spend at this location. You can move around. >> You could literally follow the money, on a map. (laughing) >> So this is your home base, you go out here, you spend this amount of money, and maybe even add more on it. So, let's say you do your grocery shopping over here, but if I moved all my business from this company to this company, how much could I save? Imagine if you could just drag and drop it and see, okay, I could actually save a couple of thousand bucks, awesome. >> And machine learning is going to totally change the game with Augmented Intelligence. AI is called Artificial Intelligence, or Augmented Intelligence, depending upon your definition. This is a good thing for consumers. >> It is, it is. >> And thinking about disruption, what do you guys, what are your thoughts on blockchain? What is your research showing? You playing around with Hyperledger at all? >> Yes we are. And blockchain, it's also quite interesting. We're doing lots of research on that. What's it's shown actually is that this is a technology that we can also use. And we can also really utilize, even the security aspects of it. If you just take that, you could really implement that. >> The identity aspect, it's federating identity around fraud, another area you can innovate on. I'm bullish on blockchain, a lot of people are skeptical, but Dave knows I really, I love blockchain. Because it's not about Bitcoin per se, it's sort of the underlying opportunity. It just seems fascinating. Dave you know, I got to get on my soapbox, blockchain soapbox. >> We've never really looked at Bitcoin as just a currency, it's move of a technology platform, and I have always been fascinated with the security angle. Virtually unhackable, put that in quotes. No need for a third party to intermediate. So many positive fundamentals, now it's guys like you figuring out, okay the practitioner saying, here's how we're going to implement it and commercialize it. >> And actually it fits in quite well with things like GDPR. This is also about opening up, the same with PSD 2. Exposing the customer data, making it available for the general public. And ultimately the goal is, so you as a consumer, me as a consumer, we own our data. >> Nadeem, thank you so much for coming on the CUBE and sharing your practitioner situation, and your advice, as well as commentary. I'll give you the last word. As you and your team embark from DataWorks 2017 and head back to the ranch, so to speak, and bring back some stuff. What are you going to work on? What's the to do item? What are you going to sharpen the saw on and cut when you get back? >> So for us on the very, very short term, it's about taking our platform and our capabilities and move it into the real enterprise world. That is our first key milestone that we are going to go for. And, I'll tell you, we're going to go all in for that. Because, unless we do that, we're not able to really attack the core of banking, which requires this, right? Please remember that a consumer doing a transaction somewhere in the world, he cannot stand and wait for ages for something to be processed. It needs to be instantaneous. So, this is what we need to do. >> You think this event, you're armed up with product. >> Absolutely, absolutely. Lots of good insight we've gotten from this. Lots of potential, lots of networking guys and other companies that we can talk to about this. >> Also great recruiting, get some developers out there too, lot of great people. Congratulations on your success and thanks for sharing this great insight here on the CUBE, exposing the data to you live on the CUBE. Silicon Angle dot TV, I'm John Furrier, with Dave Vellante my co-host, more great coverage stay with us here live in Munich, Germany for DataWorks 2017 Summit. We'll be right back.