>> Live from Las Vegas, it's theCUBE. Covering Informatica World 2019, brought to you by Informatica. >> Welcome back everyone to theCUBE's live coverage of Informatica World here in Las Vegas. I'm your host, Rebecca Knight along with my co-host John Furrier, we are joined by Jitesh Ghai, he is the Senior Vice President and General Manager Data Quality, Security and Governance at Informatica. Thank you so much for coming or returning to the show Jitesh. >> My pleasure, happy to be here. >> So, this is a real moment for data governance, we have the anniversary of GDPR and the California Privacy Act it's a topic at Dabos, there is growing concern among the public and lawmakers over security and privacy, give us the lay of the land from your perspective. >> Right, you know it is a moment for data governance, what's exciting in the space is governance was born out of risk and compliance and managing for risk and compliance, but really what it was mandating was healthy data management practices, how do we give the regulators comfort that our data is of high quality, that we know the lineage of where data is coming from that we know how the business relies on the data what is critical data? And while it was born to give the regulators comfort, what organizations very quickly realized is well when you democratize data, you need to give everybody that comfort, you need to give your data scientists, your data analysts, that same level of contextual understanding of their data right, where did it come from? What's the quality of it? How does the business use it, rely on it? And so that has been a tremendous opportunity for us, we've supported organizations, financial services from a BCBS 239 CCAR, counterparty credit risk, but what's happened is from a data democratization, data scale perspective, self-service analytics perspective, is what moved from terabytes to petabytes. We've moved from data warehouses, to data lakes and you can't democratize data unless there's a governed framework. I don't know, it sounds kind of like wait, democratizing data is supposed to be free data everywhere, but without some governed framework, it's a bit of a mess, and so what we're enabling organizations is the effective consumption and understanding of where their data is, discovering it, so that the right people can consume the data that they care about, the right data scientists can build the right models, the right analysts can build the right reports and the executives get the right confidence on what reports they're getting, what KPI's they're getting. >> One of the things that we talked last year, you had a couple customers on, you had told a great story, you guys had had the benefit as a long-standing company, 25 years in the private for large-customer base, but the markets changed, you mentioned governance I mean we're in the one year-anniversary of GDPR. >> Right. >> And I think everyone's kind of like OK what happened last year? More privacy laws are coming and one of the themes this year is clarity with data, but also in the industry you know access to data, making data addressable, because AI needs data sets, cloud has proven that, SAS business models, using data winning formula, that's clear if you're born in the cloud. Enterprises now want that same kind of SAS-like execution on the applications side, whether it's SAS or using AI for instance, >> Right. >> So when you have more regulation, inherent nature is to oh like more complexity, how are customers dealing with the complexity of this, because they want to free it up, but at the same time they want to make sure that they can respect the laws for individuals, but also governments aren't that smart either so you know, the balance there, what's the strategy? >> And therein lies the challenges with privacy specifically, it's not just about quality counterparty credit risk in like five or seven systems in a data warehouse, it's all the data in your enterprise, it's the data in production, there's the data in your DevOps environment, it's all your data literally, structured all the way to unstructured data like Word, PDFs, Powerpoints. And you need a governing framework around it, you need to enable organizations to be able to discover where is there sensitive information, how is there sensitive information proliferating through the organization? Is it protected? Is it not protected? And what's particularly, you know, we're all consumers, I'm pretty confident some or all of our data has been breached at some point, enabling organizations, what these privacy regulations are doing is they are giving us, as individuals, rights to go to the organizations we transact with and ask them, what are you doing with out data? Forget my data or at least tell me how you're processing it and get my consent for the data. >> Yeah, I mean policy and business models are certainly driving that and with regulation, I see that, but the question is that when you move the impact to the enterprise, you got storage drives. You store it on drives as a storage administrator you've got software abstractions with data, like you guys do. So, it's complicated, so the question is, for you, is what are customers doing now? What's the answer to all this? >> The answer really comes down to you need to scale to the scope of the problem, it's a thousand x-increase, you're going from terabytes to petabytes right? And so, you need an AI, an ML, an intelligent solution that can discover all of this information, but it can map it to John Furrier, this is where John Furrier's information is, it's in the human capital management system, the CRM system, organizations know, may start knowing whether sensitive data is, but hey don't know who it belongs to, so when you go to invoke your right to be forgotten or portability, today, what we're enabling organizations with is hey, we'll help you discover the sensitive information, but we'll also tell you who it belongs to, so that when John shows up or Rebecca, you show up, you just have to punch in their name and we'll tell you all the systems, that it's in. That is something that requires teams of database administrators, lawyers, system administrators that needs to be automated, to truly realize the potential of these privacy regulations, while enabling organizations to continue to innovate and disrupt with data. >> What's your take on whether or not consumers truly understand the scope of these privacy regulations, I mean talking about GDPR and you get the pop-ups that say do you consent and you just say yes, I just need to get to this site and so you blithely, just press yes, yes, yes so you are technically giving your consent, but do you, I mean what's your take, do consumers truly understand what they're doing here? >> You know, I think historically, we've all said yes, yes, yes, over the last, I would say two years with growing regulations and significant breaches, there is a change in customer expectations, you know, there's a stat out there in the event of a data breach, two-thirds of consumers of a particular organization blame the organization for the breach, not the hackers, right, so it's a mindshift in all of us, where you're the custodian of my data I'm counting on you, whatever organization I'm transacting with ,to ensure and preserve my privacy, ensure my data's protected. So, that's a big shift that's happened, so whether you're doing it for regulatory reasons, CCPA North America, there's several other state-wide regulations coming out or GDPR, the consumer expectation, forget regulations, it's brand preservation, it's customer trust, it's customer experience, that organizations are really having to solve for from a privacy standpoint. >> Tell what the news around yesterday around the shift of the trust pieces, because that's a huge deal. Because trust is shifting, expectations are shifting, so when you have shifting expectations, with users and buyers, customers, the experience has to shift. So, take us through what's the new things? >> Well, the new things are, you know, you look at we're enabling organizations to be data-driven, we're enabling organizations to transform, build new products, new services, be more efficient and for that, you need to enable them to get access to data. The counter, the tension on the other end is how do we get them broad-based access while ensuring privacy, right, and that's the balance. How do we enable them to be customer-centric and optimal in engaging with their customers while preserving the privacy of their customers and that really comes down to having a detailed understanding of what your critical data is, where is it in the organization and how an organization is using that data. Enabling an organization to know that they're processing data with the appropriate consent. >> What's interesting to me, when I was with press yesterday, is also the addition of how the cloud players are coming onboard, because you know, one constituent that's not mentioned in that statement is that you guys are kind of keeping an eye on, that are impacted by this, is developers, because you know developers like infrastructures coded with DevOps. Don't want to be provisioning networks and storage, they just write to the API's. Data is kind of going through that similar experience where, if I'm a developer doing an IOT app, I'm just going to use the cloud. I put the data there, I don't need to have a mismatch of mechanisms to deal with some governance compliance rules. >> Correct and that's why it needs to be built-in by design. And you know there's this connotation that- >> Explain that, what does built-in by design mean. >> Well you need to have privacy built-into how you as a business operate, how you as a DevOps team or development team, build products, if that's built-in to how you operate, you enable the innovation without falling into the pitfalls of oh you know what we broke some privacy regulations there we breached our customers trust there, we used data or engaged with them in manner that they weren't comfortable with. >> So, don't retro-fit after the fact? Think holistically on the front-end of the transformation in architecture. >> It's an enabler, in that if you do it right to begin with, you can continue to innovate and engage effectively, versus bolting it on as an afterthought and retro-fitting. >> It really seems like it is this evolution in thinking from this risk and compliance, overdoing this to check all the boxes, versus here are our constraints, but our constraints are actually liberating, is what you're saying. >> Right, but you can't democratize data, without giving the consumers of that data an understanding of the quality of that data, the trustworthiness of that data, the relevance of the data to the business, you give them that and now you're enabling your analytics, your data scientists, your analytics organizations to innovate with that data with confidence and if you do it within a framework of privacy, you're ensuring that you're preserving customer trust while you're automating and building intelligent and engaging customer experiences. >> What I love about the data business right now, is it's exciting because it's real specific examples of impact, security, you know, national security, to hackers, to just general security, privacy of the laws, But, I've seen the development angles interesting too, so when you got these two things moving, customers can ignore this, it's not like back-up and recovery where same kind of ethos is there, you don't want to think about it after the fact, you want to build it in, you know, there's certainly reasons why you do that, in case there's a disaster, but data is highly impactful all the time. This is a challenge, you guys can pull this off. >> Well you know, it's a, with privacy, it's no longer about a few systems, it's all your data and so the scope is the challenge and the scale applies for privacy, the scale applies for making data available enterprise wide and that's where you need and you know we spoke about AI needs data, well data also needs AI. And that's where we're leveraging AI and ML. Building out intelligence, to help organizations solve that problem and not do it manually. >> You know, I've said it on theCUBE, you've probably heard it many times, I say it all the time, scale is the new competitive advantage. Value is the new lock-in. No proprietary software anymore, but technology is needed. I want to ask you, you've been talking about this with some of your customers last year around data is that you need more scale, because AI needs more access to data, because the more visibility into data, the smarter, machine learning and AI applications can become. So Scale is real. What is the, what are you, you guys have some scalosity in your customers, you got the end-to-end, got the catalog and everything is kind of looking good, but you have competition How would you compare to the competition, when people say hey Jitesh, a start-up just popped out or XYZ company's got the solution, why should I go with them or you? What's the difference, what's the competitive angle? >> You know, the way we're thinking the problem is founded on governance is an enabler it's not about locking things down for risk and compliance, because you know, the regulators want to know that this particular warehouse is highly tightly controlled, it's about getting the data out there, it's about enabling end-users to have a contextual understanding when you're doing that for all of your data, within around, that's a thousand X-increase in the data, it's a thousand X-increase in your constituents, you're not supporting, the risk and compliance portions of the organization, you're supporting marketing, you're supporting sales, you're supporting business operations, supply chain, customer-onboarding and so with the problem of scale, practices of the past, which were typically manual laborious, but hey at the risk of non-compliance, we just had to deal with them, don't practically in any way scale, to the requirements of the future which is a thousand X-increase in consumers and that's where intelligence and AI and ML come in. >> The question I have for you is, where should customers store their data? Is there an answer to that on premises or in the cloud? What are they doing? >> The answer is yes, (Knight laughs) the customer should store their data, what we see, the world is going to be hybrid, mainframes are still here, on-premise will still be here many years from now. >> So you're taking the middle of the road here, so >> There's Switzerland. >> You're saying whatever they want on-premise or cloud, is there a preference you see with customers? >> Well, you know it depends on the applications , depends on regulations, historically regulations especially in financial services, have mandated a more on-premise stance, but those regulations, are also evolving and so we see, the global investment banks all of a sudden, we're having all sorts of conversations about enabling them to move select portions of their data estate to the cloud, enabling them to be more agile, so the answer is yes and it will be for a very long time to come. >> Final question, one of the most pressing problems in the technology industry is the skills gap. I want to hear your thoughts on it, how as a Senior Executive at Informatica, how worried are you about finding qualified candidates for your open-roles? >> You know, it is a challenge, good news is, we're a global organization, my teams are globally-distributed. I have teams in Europe, North America and Asia and the good part about that is if you can't find it in the valley, you can certainly find the talent elsewhere, and so while, it is a challenge, we're able to find talented engineers, software developers, data scientists, to help us innovate and build the intelligence capabilities to solve the productivity challenges, the scale challenges of data consumption. >> Jitesh, talk about the skills required for people coming out of school, take your Informatica hat off, put your expertise hat on, data guru hat, knowing that data is going to continue to grow, continue to have more impact across the board, from coding to society affix, whatever, what are some of the key skills in training, classes or courses or areas of expertise that people an dial-up or dig into that might be beneficial to them that may or may not be on the radar curriculum or, say is, part of school curriculum, >> you know we engage with universities in North America, in Europe, in Asia, we have a large development center in India and we're constantly, engaging with them. We're on various boards at various universities, advisory standpoint, big data standpoint and what we're seeing is as we engage with these organizations, we're able to feed back on where the market is going, what the requirements are, the nature of data science, the enabling technologies such as platforms like Spark, languages like Python and so we're working with these schools to share our perspectives, they in turn, are incorporating this into their curriculums and how they train future data scientists. >> When you see a young gun out there that's kicking butt and taking names and data, what are some of the backgrounds? Is it math, is it philosophy, is there a certain kind of pattern that you've seen as the makeup of just the killer data person? >> You know, it's interesting, you mention philosophy, I'm a big, I've hired many philosophy majors that have been some of the best architects, having said that, from a data science perspective, it's all about stats, it's all about math and while that's an important skillset to have, we're also focused on making their lives easier, they're spending 70% of their time, doing data engineering versus data science and so while they are being educated from a stats, from a data science foundation, when they come into the industry, they end up spend 70% of their time doing data engineering, that's where we're helping them as well. >> So study your Socrates and study your stats. >> I like that. (Knight and Furrier laugh) >> Jitesh, thank you so much for coming on theCUBE. >> My pleasure, happy to be here, thank you. >> I'm Rebecca Knight for John Furrier, you are watching theCUBE.

>> Announcer: From Berlin, Germany, it's the CUBE covering Dataworks Summit Europe 2018. Brought to you by Hortonworks. >> Welcome to the CUBE, we're here at Dataworks Summit 2018 in Berlin. I'm James Kobielus. I am the lead analyst for big data analytics on the Wikibon team of SiliconANGLE Media On the CUBE, we extract the signal from the noise and here at Dataworks Summit, the signal is big data analytics and increasingly the imperative for many enterprises is compliance with GDPR, the General Data Protection Regulation comes in five weeks, May 25th. There's more things going on so what I'm going to be doing today for the next 20 minutes or so is from Hortonworks I have Abhas Ricky who is the director of strategy and innovation. He helps customers, and he'll explain what he does, but at a high level, he helps customers to identify the value of investments in big data, analytics, big data platforms in their business. And Abhas, how do you justify the value of compliance with GDPR. I guess, the value would be avoid penalties for noncompliance, right? Can you do it as an upside as well? Is there an upside in terms of if you make an investment, and you probably will need to make an investment to comply, Can you turn this around as a strategic asset, possibly? Yeah, so I'll take a step back first. >> James: Like a big data catalog and so forth. >> Yeah, so if you look at the value part which you said, it's interesting that you mentioned it. So there's a study which was done by McKinsey which said that only 15% of executives can understand what is the value of a digital initiative, let alone big data initiative. >> James: Yeah. >> Similarly, Gardner says that if you look at the various portraits and if you look at various issues, the fundamental thing which executives struggle with identifying the value which they will get. So that is where I pitch in. That is where I come in and do a data perspective. Now if you look at GDPR specifically, one of the things that we believe, and I've done multiple blogs around that and webinars, GDPR should be treated at a business opportunity because of the fact that -- >> James: Any opportunity? Business opportunity. It shouldn't necessarily be seen as a compliance burden on costs or your balance sheets because of the fact, it is the one single opportunity which allows you to clean up your data supply chain. It allows you to look at your data assets with a holistic view, and if you create a transparent data supply chain, and your IT systems talk to each other. So some of the provisions, as you know, in addition to right to content, right to portability, etc. It is also privacy by design which says that you have to be proactive in defining your IT systems and architecture. It's not necessarily reactive. But guess what? If you're able to do that, you will see the benefits in other use cases like single view of customer or fraud or anti-money laundering because at the end of the day, all GDPR is allowing you to say is that where do you store your data, what's the lineage, what's the provenance? Can you identify what the personally identifiable information is for any particular customer? And can you use that to your effect as you go forward? So it's a great opportunity because to be able to comply with the provisions, you've got to take steps before that which is essentially streamlining your data operations which obviously will have a domino effect on the efficiency of other use cases. So I believe it's a business opportunity. >> Right, now part of that opportunity in terms of getting your arms around what data you have, when the GDPR is concerned, the customer has a right to withhold consent for you and the enterprise that holds that data to use that personal data of theirs which they own for various and sundry reasons. Many enterprises and many of Hortonworks customers are using their big data for things like AI and machine learning. Won't this compliance with GDPR limit their ability to seize the opportunity to build deep learning and so forth? What are customers saying about that? Is that going to be kind of a downer or a chilling effect on their investments in AI and so forth? >> So there's two elements around it. The first thing which you said, there are customers, there's machine learning in AI, yes, there are. But broadly speaking, before you're able to do machine learning and AI, you need to get your data sets onto a particular platform in a particular fashion, clean data, otherwise, you can't do AI or machine learning on top of it. >> James: Right. So the reason why I say it's an opportunity is that because you're being forced by compliance to get that data from every other place onto this platform. So obviously those capabilities will get enhanced. Having said, I do agree if I'm an organization which does targeting, retargeting of customers based on multiple segmentations and then one of the things is online advertisements. In that case, yes, your ability might get affected, but I don't think you'll get prohibited. And that affected time span will be only small because you just adapt. So the good thing about machine learning and AI is that you don't create rules, you don't create manual rules. They pick up the rules based on the patterns and how the data sets have been performing. So obviously once you have created those structures in place, initially, yes, you'll have to make an investment to alter your programs of work. However, going forward, it will be even better. Because guess what? You just cleaned your entire data supply chain. So that's how I would see that, yes, a lot of companies, ecommerce you do targeting and retargeting based on the customer DNA, based on their shopping profiles, based on their shopping ad libs and then based off that, you give them the next best offer or whatever. So, yes, that might get affected initially, but that's not because GDPR is there or not. That's just because you're changing your program software. You're changing the fundamental way by which you're sourcing the data, the way they are coming from and which data can you use. But once you have tags against each of those attributes, once you have access controls, once you know exactly which customer attributes you can touch and you cannot for the purposes, do you have consent or not, your life's even better. The AI tools or the machine learning algorithms will learn from themselves. >> Right, so essentially, once you have a tight ship in terms of managing your data in line with the GDPR strictures and so forth, it sounds like what you're saying is that it gives you as an enterprise the confidence and assurance that if you want to use that data and need to use that data, you know exactly how you've the processes in place to gain the necessary consents from customers. So there won't be any nasty surprises later on of customers complaining because you've got legal procedures for getting the consent and that's great. You know, one of the things, Abhas, we're hearing right now in terms of compliance requirements that are coming along, maybe not apart of GDPR directly yet, but related to it is the whole notion of algorithmic transparency. As you build machine learning models and these machine learning models are driven into working applications, being able to transparently identify if those models make, in particular, let's say autonomous action based on particular data and particular variables, and then there is some nasty consequences like crashing an autonomous vehicle, the ability, they call it explicably AI to roll that back and determine who's liable for that event. Does Hortonworks have any capability within your portfolio to enable more transparency into the algorithmic underpinnings of a given decision? Is that something that you enable in your solutions or that your partner IBM enables through DSX and so forth? Give us a sense whether that's a capability currently that you guys offer and whether that's something in terms of your understand, are customers asking for that yet or is that too futuristic? >> So I would say that it's a two-part question. >> James: Yeah. >> The first one, yes, there are multiple regulations coming in, like Vilica Financial Markets, there's Mid Fair, the BCBS, etc. and organizations have to comply. You've got the IFRS which span to brokers, the insurance, etc., etc. So, yes, a lot of organizations across industries are getting affected by compliance use cases. Where does Hortonworks come into the picture is to be able to be compliant from a data standpoint, A you need to be able to identify which of those data sources you need to implement a particular use case. B you need to get them to a certain point whereby you can do analytics on that And then there's the whole storage and processing and all of that. But also which you might have heard at the keynote today, from a cloud perspective, it's starting to get more and more complex because everyone's moving to the cloud which means, if you look at any large multi-national organization, most of them have a hybrid cloud structure because they work with two or three cloud vendors which makes the process even more complex because now you have multiple clusters, you have have on premise and you have multiple different IT systems who need to talk to each other. Which is where the Hortonworks data plan services come into the picture because it gives you a unified view of your global data assets. >> James: Yes. >> Think of it like a single pane of glass which whereby you can do security and governance across all data assets. So from those angles, yes, we definitely enable those use cases which will help with compliance. >> Making the case to the customer for a big data catalog along the lines of what you guys offer, in making the case, there's a lot of upfront data architectural work that needs to be done to get all you data assets into shape within the context of the catalog. How do they justify making that expense in terms of hiring the people, the data architects and so forth needed to put it all in shape. I mean, how long does it take before you can really stand up in your working data catalog in most companies? >> So again, you've asked two questions. First of all is how do they justify it? Which is where we say that the platform is a means to an end. It's enabling you to deliver use cases. So I look at it in terms of five key value drivers. Either it's a risk reduction or it's a cost reduction or it's a cost avoidance. >> James: Okay. >> Or it's a revenue optimization, or it's time to market. Against each one of these value drivers, or multiple of them or a combination of them, each of the use cases that you're delivering on the platform will lead you to benefits around that. My job, obviously, is to work with the customers and executes to understand what will that be to quantify the potential impact which will then form the basis and give my customer champions enough ammunition so that they can go back and justify those investments. >> James: Abhas, we're going to have to cut it short, but I'm going to let you finish your point here, but we have to end this segment so go ahead. >> That's fine. >> Okay, well, anyway, have had Abhas Ricky who is the director of strategy and innovation at Hortonworks. We're here at Dataworks Summit Berlin. And thank you very much Sorry to cut it short, but we have to move to the next guest. >> No worries, pleasure, thank you very much. >> Take care, have a good one. >> Thanks a lot, yes. (upbeat music)

>> narrator: From Las Vegas it's theCUBE covering IBM Think 2018. Brought to you by IBM. >> We're back at IBM Think 2018. My name is Dave Vellante. I'm here with Peter Burris, my co-host, and you're watching theCUBE, the leader in live tech coverage. Our day two of our wall to wall coverage of IBM Think. Madhu Kochar is here. She's the Vice President of Analytics, Product Development at IBM and she's joined by Pat Maqetuka >> Close enough. She's a data officer at Nedbank. Ladies, welcome to theCUBE. You have to say your last name for me. >> Patricia Maqetuka. >> Oh, you didn't click >> I did! >> Do it again. >> Maqetuka. >> Amazing. I wish I could speak that language. Well, welcome. >> Madhu & Pat: Thank you. >> Good to see you again. >> Madhu: Thank you. >> Let's start with IBM Think. New show for you guys you consolidated, you know, six big tent events into one. There's a lot of people, there's too many people, to count I've been joking. 30, 40 thousand people, we're not quite sure, but how's the event going for you? What are clients telling you? >> Yeah, no, I mean, to your point, yes, we brought in all three big pillars together; a lot of folks here. From data and analytics perspective, an amazing, amazing event for us. Highlights from yesterday with Arvind Krishna on our research. What's happening. You know, five for five, that was really inspiring for all of us. You know, looking into the future, and it's not all about technology, was all about how we are here to help protect the world and change the world. So that as a, as a gige, as an engineer, that was just so inspiring. And as I was talking to our clients, they walk away with IBM as really a solution provider and helping, so that was really good. I think today's, Ginni's, keynote was very inspiring, as well. From our clients, we got some of our key clients, you know, Nedbank is here with us, and we've been talking a lot about our future, our strategy. We just announced, Ginni actually announced, our new product, IBM Cloud Private for Data. Everything around data, you know, where we are really bringing the power of data and analytics all together on a private cloud. So that's a huge announcement for us, and we've been talking a lot with our clients and the strategies resonating and particularly, where I come from in terms of the co-ordinance and integration space, this is definitely becoming now the "wow" factor because it helps stitch the entire solutions together and provide, you know, better insights to the data. >> Pat from your perspective, you're coming from Johannesburg so you probably like the fact that there's all IBM in one, so you don't have to come back to three or four conferences every year, but love your perspectives on that, and can you please tell us about Nedbank and your role as Chief Data Officer. >> Nedbank is one of the big five financial banks in South Africa. I've been appointed as the CDO about 18 months ago, so it's a new role in the bank per say. However, we're going through tremendous transformation in the bank and especially our IT eco-system has been transformed because we need to keep up with what is happening in the IT world. >> Are you the banks first Chief Data Officer? >> Definitely yes. I'm the first. >> Okay. So, you're a pioneer. I have to ask you then, so where did you start when you took over as the Chief Data Officer? I mean banking is one of those industries that tends to be more Chief Data Officer oriented, but it's a new role, so where did you start? >> Well we are not necessary new in the data, per say. We have had traditional data warehousing functions in the organization with traditional warehousing or data roles in the organization. However, the Chief Data role was never existent in the bank and actual fact, the bank appointed two new roles 18 months ago. One of them was the Chief Digital Officer, who is my colleague, and myself being appointed as the Chief Data Officer. >> Interesting, okay. I talked to somebody from Northern Trust yesterday and she was the lead data person and she said, "I had to start with a mission. We had to find the mission first and then we looked at the team, and then we evolved into, how we contribute to the business, how we improve data quality, who has access, what skills we needed." Does that seem like a logical progression and did you take a similar path? >> I think every bank will look at it differently or every institution. However, from a Nedbank perspective, we were given the gift by the regulator in bringing the BCBS 239 compliancy into play, so what the bank then did, how do we leverage, not just being compliant, but leveraging the data to create competitive advantage, and to create new sources of revenue. >> Okay. Let's talk about, Madhu we talked about this in New York City, you know, governance, compliance, kind of an evil word to a lot of business people. Although, your contention was "Look, it's reality. You can actually turn it into a positive." So talk about that a little bit and then we can tie it into Nedbank's experiences. >> Yeah, so I firmly believe in, you know, in the past governance 1.0 was all about compliance and regulations, very critical, but that's all we drew. I believe now, it's all about governance 2.0, where it's not just the compliance, but how do I drive insights, you know, so data is so, so critical from that perspective, and driving insights quicker to your businesses, is going to be very important, so as we engage with Nedbank and other clients as well, they are turning that because they are incumbents. They know their data, they've got a lot of data, you know, some of, they know sitting in structure, law structure land, and it's really, really important that they quickly able to assess what's in it, classify it, right, and then quickly deliver the results to the businesses, which they're looking for, so we're, I believe in lieu of governance 2.0, and compliance and regulations are always going to be with us, and we're making, actually, a lot of improvements in our technology, introducing machine learning, how we can do these things faster and quicker. >> So one of the first modern pieces of work that Peter and I did was around data classification and that seems to be, I heard this theme before, it seems to be a component or a benefit of putting governance in place. That you can automate data classification and use it to affect policy, but Pat, from your standpoint, how do you approach governance, what are the business benefits beyond "we have to do this"? >> Like I earlier on alluded to, we took the regulation as a gift and said, "How do we turn this regulation into benefits for the organization?" So in looking at the regulation we then said, "How do we then structure the approach?" So we looked at the two prompts. The first was, the right to win. The right to win meaning that we are able to utilize the right to compete approach from a regulation perspective, to create a platform and a foundation for analytics for our organization. We also created the blueprint for our enterprise data program and in the blueprint, we also came up with key nine principles of what it means to stay true to our data. I.e. you mentioned classification, you mentioned data politic, you mentioned lineage. Those are the key aspects within our principles. The other key principle we also indicated was the issue around duplication. How do we ensure that we describe data once, we ingest it once, but we use it multiple times to answer different questions, and as you are aware, in analytics, the more you mine the data, the more inquisitive you become, so it is, (clears throat) Sorry. It's not been from data to information, information to insight, and eventually insights to foresight, so looking into the future, and now you bring it back into data. >> And also some points that you've made Pat, so the concept is, one of the challenges of using the fuel example, is that governance of fuel, is still governance of a thing. You can apply it here, you can apply it there, you can't apply it to both places. Data's different and you were very, very accurate when you said "We wanted to find it once, we want to ingest it once, we want to use it multiple times". That places a very different set of conditions on the types of governance and in many respects, in the past, other types of assets where there is this sense of scarcity, it is a problem, but one of the things that I'm, and this is a question, is the opportunity, you said the regulatory opportunity, is the opportunity, because data can be shared, should we start treating governance really as a way of thinking about how to generate value out of data, and not a way of writing down the constraints of how we use it. What do you think about that? >> I think you are quite right with that because the more you give the people the opportunity to go and explore, so you unleash empowerment, you unleash freedom for them to go and explore. They will not see governance as a stick like I initially indicated, but they see it as business as usual, so it will come natural. However, it doesn't happen overnight. People need to be matured, organization is to be matured. Now, the first step you have to do is to create those policies, create awareness around the policies, and make sure that the people who are utilizing the data are trained in to what are the do's and the don'ts. We are fully aware that cyber security's one of our biggest threats, so you can also not look at how you create security around your data. People knowing that how I use my data it is an asset of the bank and not an asset of an individual. >> I know you guys have to go across the street, but I wanted to get this in. You're a global analytics global elite client; I want to understand what the relationship is. I mean, IBM, why IBM, maybe make a few comments about your relationship with the company. >> I think we as Nedbank, we are privileged, actually, to be inculcated into this global elite program of IBM. That has helped me in actual, in advancing what we need to do from a data perspective because anytime I can pick up a phone to collaborate with the IBM MaaS, I can pick up the phone whenever I need support, I need guidance. I don't have to struggle alone because they've done it with all the other clients before, so why should I reinvent the wheel, whereas someone else has done it, so let me tap into that, so that that can progress quicker than try it first. >> Alright. Madhu, we'll give you the final word. On Think and your business and your priorities. >> So, Think is amazing, you know, the opportunity to meet with all our clients and coming from product development, talking about our strategy and getting that validation is just good, you know, sharing open road maps with clients like Nedbank and our other global elites, you know. It gives us an opportunity, not just sharing of the road maps, but actually a lot of co-creation, right, to take us into the future, so I'm having a blast. I got to go run over and meet a few other clients, but thank you for having us over here. It's a pleasure. >> You're very welcome and thank you so much for coming on and telling your story, Pat, and Madhu, always a pleasure to see you. >> Thank you. >> Alright, got to get in your high horse and go. Thanks for watching everybody, we'll be right back after this short break. You're watching theCUBE live from IBM Think 2018. We'll be right back. (electronic music)

>> Announcer: Live from San Francisco, it's The Cube covering Informatica World 2017. Brought to you by Informatica. >> Okay, welcome back everyone. We are here live in San Francisco for The Cube's exclusive coverage of Informatica World 2017. I'm John Furrier, this is Siliconangle's flagship program, we go out to the events and (he mumbles). My next guest is Jitesh Ghai who's the Vice President General Manager of data quality and governance for Informatica. Welcome to The Cube, thanks for joining us today. >> Happy to be here, John. Pleasure. >> So, two things right out of the gate. One, data quality and governance, two of the hottest topics in the industry, never mind within Informatica. You guys are announcing a lot of stuff, customers are pretty happy, you got a solid customer base. >> That's right. >> Product's been blooming, you got a big brand behind you now. This is important. There's laws now in place coming online in 2018, I think it's the GDPR. >> That's right. >> And there's a variety of other things, but more importantly customers got to get hold of their data. >> That's right. >> What's your take and what are you announcing here at the show? >> Well, you know, from a data governance and compliance and overall quality standpoint, data governance started off as a stick, a threat of regulatory pressure, but really the heart of what it is is effective access to and consumption of data, trusted data. And through that exercise of the threat of a stick, healthy practices have been implemented and that's resulted in an appreciation for data governance as a carrot, as an opportunity to innovate, innovate with your data to develop new business models. The challenge is as this maturation in the practice of data governance has happened there's been a realization that there's a lot of manual work, there's a lot of collaboration that's required across a cross-functional matrixed organization of stakeholders. And there's the concept of ... >> There's some dogma too, let's just face it, within organizations. I got all this data, I did it this way before. >> Right. >> And now, whoa, the pressure's on to make data work, right, I mean that's the big thing. >> That's exactly right. So, you collaborate, you align, and you agree on what data matters and how you govern it. But then you ultimately have to stop documenting your policies but actually make it real, implement it, and that's where the underlying data management stack comes into place. That could be making it real for regulatory, financial regulations, like BCBS 239 and CCAR, where data quality is essential. It could be making it real for security related regulations where protection is essential, like GDPR, the data protection regulation in the EU. And that's where, Informatica is launching a holistic enterprise data governance offering that enables you to not just document it, or as one CDO said to me, "You know, at some point you've got to stop talking about it, "you actually have to do it." To connecting the conceptual, the policies, with the underlying physical systems, which is where intelligent automation with the underlying data management portfolio, the industry-leading data management portfolio that we have, really delivers significant productivity benefits, it's really redefining the practice of data governance. >> Yeah, most people think of data as being one of those things, it's been kind of like, whether it's healthcare, HIPAA old models, it's always been an excuse to say no. "Whoa, we don't do it that way." Or, "Hey." It's kind of become a no-op kind of thing where, "No, we don't want to do any more than data." But you guys introduced CLAIR which is the acronym for the clairvoyant or AI, it's kind of a clever way to brand. >> That's right. >> That's going to bring in machine learning augmented intelligence and cool things. That only, to me, feels like you're speeding things up. >> That's exactly right. >> When in reality governance is more of a slowdown, so how do you blend the innovation strategy of making data freely available ... >> Right. >> ..and yet managing the control layer of governance, because governance wants to go slow, CLAIR wants to go fast, you know. Help me explain that. >> Well, in short, sometimes you have to go slow to go fast. And that's the heart of what our automated intelligence that CLAIR provides in the practice of data governance, is to ensure that people are getting access to, efficient access to trusted data and consuming it in the right context. And that's where you can set, you can define a set of policies, but ultimately you need those policies to connect to the right data assets within the enterprise. And to do that you need to be able to scan an entire enterprise's data sets to understand where all the data is and understand what that data is. >> Talk about the silver bullet that everyone just wants to buy, the answer to the test, which is ungettable, by the way, I believe, we just had Allegis on, one of your customers, and their differentiation to their competition is that they're using data as an asset but they're not going all algorithmic. There's the human data relationship. >> Absolutely. >> So there's really no silver bullet in data. You could use algorithms like machine learning to speed things up and work on things that are repeatal tasks. >> Right. >> Talk about that dynamic because governance can be accelerated with machine learning, I would imagine, right? >> Absolutely, absolutely. Governance is a practice of ensuring an understanding across people, processes and systems. And to do that you need to collaborate and define who are the people, what are your processes, and what are the systems that are most critical to you. Once you've defined that it's, well, how do we connect that to the underlying data assets that matter, and that's where machine learning really helps. Machine learning tells you that if you define customer id as a critical data element, through machine learning, through CLAIR, we are able to surface up everywhere in your organization where customer id resides. It could be cmd id, it could be customer_id, could be customer space id, cust id. Those are all the inferences we can make, the relationships we can make, and surface all of that up so that people have a clear understanding of where all these data assets reside. >> Jitesh, let's take a step back. I want to get your thoughts on this, I really want you to take a minute to explain something for the folks watching. So, there's a couple of different use cases, at least I've observed in a row and the wikibon team has certainly observed. Some people have an older definition of governance. >> Right. >> What's the current definition from your standpoint? What should people know about governance today that's different than just last year or even a few years ago, what's the new picture, what's the new narrative for governance and the impact to business? >> You know, it's a great question. I held a CDO summit in February, we had about 20 Chief Data Officers in New York and I just held an informal survey. "Who implements data governance programs "for regulatory reasons?" Everybody put their hand up. >> Yeah. >> And then I followed that up with, "Who implements data governance programs "to positively affect the top line?" and everybody put their hand up. That's the big transition that's happened in the industry is a realization that data governance is not just about compliance, it's also about effective policies to better understand your data, work with your data, and innovate with your data. Develop new business models, support your business in developing those new business models so that you can positively affect the top line. >> Another question we get up on The Cube all the time, and we also observe, and we've heard this here from other folks at Informatica and your customers have said, getting to know what you actually have is the first step. >> Right. >> Which sounds counter-intuitive but the reality is that a lot of folks realize there's an asset opportunity, they raise their, hey, top line revenue. I mean, who's not going to raise their hand on that one, right, you get fired. I mean, the reality is this train's coming down the tracks pretty fast, data as an input into value creation. >> That's exactly right. >> So now the first step is oh boy, just signed up for that, raise my hand, now what the hell do I have? >> Right. >> How do you react to that? What's your perspective on that? >> That's where you need to be able to, google indexed the internet to make it more consumable. Actually, a few search engines indexed the internet. Google came up with sophistication through its page-ranking algorithm. Similarly, we are cataloging the enterprise and through CLAIR we're making it so that the right relevant information is surfaced to the right practitioner. >> And that's the key. >> That is the key. >> Accelerating the access method, so increase the surface area of data, have the control catalog for the enterprise. >> That's right. >> Which is like your google search analogy. A little harder than searching the internet, but even google's not doing a great job these days, in my opinion, I should say that. But there's so many new data points coming in. >> That's right. >> So now the followup question is, okay, it's really hard when you start having IOT come in. >> That's right. >> Or gesture data or any kind of data coming in. How do you guys deal with that? How does that rock your world, as they say? >> And that's where effective consumption of data permeates across big data, cloud, as well as streaming data. We have implemented, in service to governance, we've implemented in-stream data quality rules to filter out the noise from the signal in sensor data coming in from aircraft subsystems, as an example. That's a means of, well, first you need to understand what are the events that matter, and that's a policy definition exercise which is a governance exercise. And then there's the implementation of filtering events in realtime so that you're only getting the signal and avoiding the noise, that's another IOT example. >> What's your big, take your Informatica hat off, put your kind of industry citizen hat on. >> Mm-hm. >> What's your view of the marketplace right now? What's the big wave that people are riding? Obviously, data, you could say data, don't say data 'cause we know that already. >> Sure. >> What should people, what do you observe out there in the marketplace that's different, that's changing very rapidly? Obviously we see Amazon stock going up like a hockey stick, obviously cloud is there. What are you getting excited about these days? >> You know, what I'm excited about is bringing broad-based access of data to the right users in the right context, and why that's exciting is because there's an appreciation that it's not the analytics that are important, it's the data that fuels those analytics that's important. 'Cause if you're not delivering trusted, accurate data it's effectively a garbage in, garbage out analytics problem. >> Hence the argument, data or algorithms, which one's more important? >> Right. >> I mean data is more important than algorithms 'cause algorithms need data. >> That's exactly right and that's even more true when you get into non-deterministic algorithms and when you get into machine learning. Your machine learning algorithm is only as good as the data you train it with. >> I mean look, machine learning is not a new thing. Unsupervised machine learning's getting better. >> Right. >> But that's really where the compute comes in, and the more data you have the more modeling you can do. These are new areas that are kind of coming online, so the question is, to you, what new exciting areas are energizing some of these old paradigms? We hear neural nets, I mean, google's just announced neural nets that teach neural nets to make machine learning easier for humans. >> Right. >> Okay. I mean, it has a little bit of computer science baseball but you're seeing machine learning now hitting mainstream. >> Right. >> What's the driver for all this? >> The driver for all this comes down to productivity and automation. It's productivity and automation in autonomous vehicles, it's productivity and automation that's now coming into smart homes, it's productivity and automation that is being introduced through data-driven transformation in the enterprise as well, right, that's the driver. >> It's so funny, one of my undergraduate computer science degrees was databases. And in the '80s it wasn't like you went out to the tub, "Hey, I'm a databaser." (He mimics uncertain mumbling) And now it's like the hottest thing, being a data guy. >> Right. >> And what's also interesting is a lot of the computer science programs have been energized by this whole software defined with cloud data because now they have unlimited, potentially, compute power. >> Right. >> What's your view on the young generation coming in as you look to hire and you look to interview people? What are some of the disciplines that are coming out of the universities and the masters programs that are different than it was even five years ago? What are some trends you're seeing in the young kids coming in, what are they gravitating towards? >> Well, you know, there's always an appreciation of, a greater appreciation for, you know, the phrase I love is, "In god we trust, all others must have data." There's an increasing growing culture around being data-driven. But from a background of young people, it's from a variety of backgrounds, of course computer science but philosophy majors, arts majors in general, all in service to the larger cause of making information more accessible, democratizing data, making it more consumable. >> I think AI, I agree, by the way, I would just add, I think AI, although it is hyped and I don't really want to burst that bubble because it's really promoting software. >> Right. >> I mean, AI's giving people a mental model of, "Oh my god, some pretty amazing things are happening." >> Sure. >> I mean, autonomous vehicles is what most people point to and say, "Hey, wow, that's pretty cool." A Tesla's much different than a classic car. I mean, you test-drive a TESLA you go, "Why am I buying BMW, Audi, Mercedes?" >> Right, exactly. >> It's a no brainer. >> Right. >> Except it's like (he mumbles), you got to get it installed. But, again, that's going to change pretty quickly. >> At this point it's becoming a table sticks exercise. If you're not innovating, if you're not applying intelligence and AI, you're not doing it right. >> Right, final question. What's your advice to your customers who are in the trenches, they raise their hand, they're committed to the mandate, they're going down the digital business transformation route, they recognize that data's the center of the value proposition, and they have to rethink and reimagine their businesses. >> Right. >> What advice do you give them in respect to how to think architecturally about data? >> Well, you know, it all starts with your data-driven transformations are only as good as the data that you're driving your transformations with. So, ensure that that's trusted data. Ensure that that's data you agree as an organization upon, not as a functional group, right. The definition of a customer in support is different from the definition of a customer in sales versus marketing. It's incredibly important to have a shared understanding, an alignment on what you are defining and what you're reporting against, because that's how you're running your business. >> So, the old schema concept, the old database world, know your types. >> Right. >> But then you got the unstructured data coming in as well, that's a tsunami IOT coming in. >> Sure, sure. >> That's going to be undefined, right? >> And the goal and the power of AI is to infer and extract metadata and meaning from this whole landscape of semi-structured and unstructured data. >> So you're of the opinion, I'm sure you're biased with being Informatica, but I'm just saying, I'm sure you're in favor of collect everything and connect the dots as you see fit. >> Well ... >> Or is that ...? >> It's a nuance, you can't collect everything but you can collect the metadata of everything. >> Metadata's important. >> Data that describes the data is what makes this achievable and doable, practically implementable. >> Jitesh Ghai here sharing the metadata, we're getting all the metadata from the industry, sharing it with you here on The Cube. I'm John Furrier here live at Informatica World 2017, exclusive Cube coverage, this is our third year. Go to siliconangle.com, check us out there, and also wikibon.com for our great research. Youtube.com/siliconangle for all the videos. More live coverage here at Informatica World in San Francisco after this short break, stay with us.