 Live from San Francisco, it's theCUBE, covering Informatica World 2016. Brought to you by Informatica. Now, here are your hosts, John Furrier and Peter Burris. Hello everyone, welcome back. We are here live in San Francisco for exclusive coverage of Informatica World. This is theCUBE SiliconANGLE Media's flagship program. We go out to the events and extract the signals of noise. I'm John Furrier, host with my co-host, Peter Burris, head of research at SiliconANGLE Media and GM of Wikibon Research. Our next guest is Jerry Held, who's a board member of Informatica, also a chairman of MemSQL, knows a little bit about databases, has been on theCUBE at Big Data SV. Welcome back, great to see you. Great to be back. Love having you on, because you're like a wealth of knowledge. You have to look back and you have to look forward. Well, look back, crescent, and then look forward. And you've been in the trenches, you've been involved in technology going way back. We've covered this last time in databases. But now, data. Well, as I recall, you were Stonebreaker's seminal graduate student. First, we started out together 45 years ago and sort of were in the delivery room of the whole relational database business. So it's been an exciting four plus decades. Things continue to change. And I think right now is, after all that time, one of the most exciting times in the whole data industry. Yeah, and we're pleased to have you. And I want to get your perspective because you have some thought leadership, not only now, but looking forward. And this abstraction layer where data now is going to be manipulated with software to provide value as key. And real time is a big part of it. And I know that's a big part of what you've been looking at right now with MemSQL and Sully Informatica. Tie those two together, this new abstraction layer, letting data fly around all over the place and be addressable and or have purpose, as Peter would say, and real time and the impact of those two things. Yeah, it's interesting you ask a question. Why is now the time for real time? We've talked about real time for eons. I mean, it's been real time systems, some of them very small real time control systems. But why is real time mainstream now? And again, you have to kind of look backward. I remember in the 70s at Tandem, we were doing online transaction processing for the first time. And we built these great databases for transaction processing. Pretty soon afterwards, we added the capability of querying and data warehousing. So being the techie guy, I went to the marketing folks and I said, I got a great idea. Why don't we build a system? We have the technical capability to put online transaction processing and analytics in the same box and sell it to our customers. And they went out and talked to the customers and every one of them said, we don't want it. We don't want it. Now, why would they say that? It's because they were so focused on getting their operational systems up and running. They had to put all their focus and priorities on getting their systems up, getting ERP and CRM and all of those things right. So getting transactions running right was important. The data warehouse was a separate system not to slow things down, look at historical data. We were talking a little bit before about the time element. It's actually easier to do the historical things, record transactions and do the predictive things, do predictive analytics and look at all this, look into the future. The hardest thing is now and that's what real time is about. Well, that wasn't the priority then. The technology wasn't there and it wasn't the business priority. But today we've got all the operational systems nailed. Yeah, maybe we're moving them to the cloud and everything but we've got them pretty well nailed. We've got good analytics systems to look back and look forward. But the big change is the ability to act on data in the moment and that's hard. Sure is, in fact, I want to build on it for a second Jerry, if I may. That the concept of real time, the only reason why the term scares me as we were talking about this particular issue is because as you said, it's been around for a long time and it suggests the technology. We just didn't have the technology to do it but we've had the technology to do it. But your customer's story going out and talking to folks about putting analytics and transaction OLTP in the same box is illustrative of something else. The business wasn't ready for it. There's an enormous cultural change associated with taking real time or data from a system, presenting it as options that then put some bounds in what people do but hopefully those are good bounds. So talk about some of that cultural impact that's happening now and is on the horizon to make it possible for people to engage systems now. Now a part of it is what's technically feasible because before you could put the systems together, they would work but you didn't want to compromise the person standing at the ATM machine who was doing an online transaction with the analyst who was doing a query. So you'd actually better off separating them. Also, the person. You mean taking the analyst out of the critical path of getting your money. Exactly, so we have one system that's doing the transaction processing and we use actually where Informatica came from, ETL. We move the data from the transaction system into the warehouse. And in the warehouse, we could have a separate system, actually reorganize the data for analysis and have two separate systems. So that was what was possible then. What's possible now, especially with in-memory systems, super high performance, you can actually ingest and analyze at the same time and same performance. And what that enables is business people to think about completely new applications. Let's just take a few examples and I'm not going to mention names here because I'm involved with enough companies. I can never tell what I can say and what I can't say, but some of these are Informatica customers, some are MemSQL customers and some are others. Real-time oil exploration. Here's a company, a big company, that has drills that are going down exploring for oil. What they did before, they took literally terabytes of data coming off of those drill bits, analyzed them for two days and figured out where to steer that drill bit. Now, with a real-time system, they're able to ingest terabytes of information and at the same time analyze, and in real-time still let steer that drill bit, saving millions of dollars in making better use of that drilling technology. If you look at all of the different applications of predictive maintenance, airlines where they're getting data right off of the engines, if you look at oil operations where they're getting information from the oil fields, you look at all these types of Internet of Things applications where data is being streamed in, you can predict and save money in real-time. We have all kinds of risk of fraud prevention. There's a lot of fraud prevention that was historical. You look at data and you see what happened and you then try to protect yourself. Now, in real-time, you can prevent fraud and stop the transaction in the moment. You don't have a lot of time. You only have the time while that car is being swiped. So we have massive opportunities across industry after industry. If you look at your Uber ride or your Lyft ride, there's geospatial database applications going on in real-time to get the right driver to you and take you to the right destination. The greatest example is last weekend I was driving to Tahoe in my new Tesla. Well, I wasn't driving. I actually only drove halfway. The car drove halfway by itself. Talk about real-time data, all kinds of data, road sensor data, existing map data, road, highway data. You have to have your hand on the wheel. Like, were you like, you know something? Well, you have to keep your hand on the wheel. You've got to kind of watch it. It's driving. And each car is sharing information with other cars through a real-time database. So when I see a pothole, the next car is going to know to avoid that pothole because the cars are communicating. It's changing every industry, real-time, in the moment, processing. You have to look at every business that you're in. There will be major changes. And this is not just one example. It's every industry. I mean, we're in the content business, so you see things like also Facebook coming out with Live Now, very emphasis on video, of course, we're doing video, but in the moment, it's very difficult. Totally agree, in all industries. But what's the tech that's impacting this? Be specific. What do you see that's the sequence? Because it's early days, we would agree with that. Talk about the tech and the impact. What is enabling it? What's holding it back, if anything? What are some of the trends and specifics you can share? Well, I think the exciting thing is there's been real-time for a long time. There have been real-time operating systems that were specialty. There have been real-time systems that have been hand-coded. Most of them have to be hand-coded because of the need for speed. The exciting thing now is there are standard technologies coming along. Tools that raise the implementer to another lever so they can get these systems implemented quickly. And let's just think about the progression that you need for a real-time system. Typically, there's some data capture. Could be from an Internet of Things device. Could be from a person or laptop, wherever it is. And you need something to capture the data. And, for instance, Informatica has some great tools that standardize the way you capture that and raise it up so that you don't have to do your own programming. The next thing you have to do is transport that data and move it to a processing place. Again, there's some old technologies that were slow. We've got some great things. In fact, in the open-source world, I think one of the best out there is Kafka, which is both a publish and subscribe or a queuing mechanism. Lots of people are using it. It allows data to be transported back and made available to lots of applications. The next thing you need to do is sort of transform that data. Just like if you were going from an OLTP system to a data warehouse, you're moving data, there's a step to clean, to protect, to transform. That's where Informatica just shines. And taking the technologies that's been developed over the last 20 years, as you hear in the conference today, Informatica is focusing on how do you take all those capabilities and raise the level up. Sure, you can go program all that in Java and spend your life doing that and having to support it all. But if you start with these diagrams where you just pluck all the pieces that have already been done, you can transform that data in real time. And part of it is just to protect it. That data coming in, if you want to analyze it in real time, I don't want you to see my credit card number, maybe the last four digits. We can program that right in at a high level using the Informatica tools. The next and really important step is the processing. So now you've processed it and you want to analyze it, they've sparked, there's all kinds of great tools to do analysis with. And the final step is storing it. Now I think one of the most interesting things about real time is that's the picture a lot of people are drawing. In fact, a lot of people draw a picture with what they call a Lambda architecture. And if you're familiar with that, but it goes in two directions, real time and then there's a batch on the side because the real time really isn't good enough. To me, Lambda, eventually when we look back at it, it's beautiful today. I think it'll be lame in the future. That's what the L will stand for. Because why do that? If you could put- Is it a slowness or is it just- It's because there hasn't been a storage engine in the past that was capable of ingesting and analyzing at the same time. That's where MemSQL comes in. Instead of analyzing and then storing, why not stream it in and while you're streaming it, analyze it. If you can analyze and ingest at the same time, you don't need a dual path. You don't need to do things twice. You do it once. You make it available at a high level with SQL, with programming languages to Spark, and you have a very flexible system. So the basic message is, technology is changing dramatically right now. It's enabling people to put real time systems up with tools that they've never had before. And the speed of processing, in-memory processing is enabling these applications like never before. So it's important that everybody in every industry look at, where's the impact of real time going to be and how can I implement it in a cost-effective way? Partly that comes back to, again, some of the cultural changes we mentioned earlier, that one of the challenges will be, does the business value data as it should? The fact that the data can be made available and can be processed and moved around with significant speeds. There is still a challenge of getting businesses to acknowledge the value of data and the role that it plays in the performing a very, very complex work. What do you see thinkers such as yourself, thought leaders, how are they describing the value of data to the business so that the business appropriately invests in it? What's the big message here at Informatical World is this data 3.0. Lots of people are talking about data as the most important asset in a business. And people are putting chief data officers in place, but get a chief data officer on your show sometime and ask him the question, if I go ask a CFO, where are all your assets, your financial assets? Don't tell me down at the penny where everything is. Except for. But now ask the CDO, the chief data officer, where are all your data assets? Who has access to all your data assets? They don't know. And the CFO and the CDO certainly will not agree. They won't. So here is the big issue I think to your question is, how do you treat data as maybe your most important asset? This is what Informatical is the intelligent data platform. It's not about your Oracle database, your SAP application, your Hadoop cluster, your NoSQL this or your MM SQL that. There's so much data in your organization. The biggest problem is coming up above all these schemas and non-schemas and whatever and having an intelligent data platform that lets you know where data is, that enables, democratizes data so that people that need to get at can get at it wherever it is. Not only on premise, but in the cloud. So data exists in all these places and a huge, huge problem is to know where it is, who has access to it to keep the right people to have access to it. And that's the main message here. I mean Informatical is doing a lot of wonderful things, but I think the most strategic is this intelligent data platform that gives control to the Chief Data Officer, the IT folks and access democratization to the many, many users within the corporation at the right level of access. I'd love to get your thoughts on that innovation layer because the layering that out gives opportunities for developers. You've been involved in tech and you've been involved in creating value. But now as you look at from today, now, and also the future, what is the value extraction from a developer's name? It's someone who wants to create value because the innovation is coming not only on the technology side, and certainly open source and the combination of software and open source will accelerate that. The business innovation, business model innovation, also technical innovation. What's the mindset of the developer? So the people that are sitting in their chairs today going, hmm, I could create value from my company. What's the mindset? What should they be thinking about? What's your advice? Share some insight into that paradigm. I think it's very important not to focus so much on the technology first, but on the outcomes first. What problem are we trying to solve? And then when you go to think about it, it's typically data centric. So what data do I need? If we can have this kind of platform where an individual can say, here's all the data available. They can start thinking in business terms about how can I reduce cost or increase revenue or make our business more efficient? But if they're thinking about bits and bytes and they're down in the weeds, they're going to miss the big innovations. It's getting that first thought at the high level and then great, it's fun to get down in the weeds and find the data and get the insights. But keep your sights high first. And if you can get that visibility about all the data available, not only inside the business, but data coming from all sorts of outside sources, you're going to be innovative. You're going to come up with good solutions. So let's talk about that for a second and bring it back to the real time concept because I think it's a great question, John. And when or what does, let me step back. When a business person thinks about making an investment or making a decision, they think in terms of what resources do I need? What money do I need? What people do I need? Where am I going to need it, et cetera? They rarely think about what data are they going to need. And it sounds like this whole motion of real time, we have to introduce new ways of thinking about modeling, new ways of thinking about business design, new ways of a number of other new ways that explicitly ask the question, what data will I need to get this done? Because you're right, it starts with the problem and then it goes to all these other things, but that one crucial issue of and what data do I need gets lost. Do you agree and do you see some, if you do, do you see some progress being made in how we think about the relationship between business problems and data starting to mature? I think it's definitely coming to the fore and people are understanding how we talk about big data. I think here we try to talk about big data, the big value. A lot of the big data was, we have a lot of data, let's go in and see what we can discover about it. So I think a lot of the big data, the initial forays into big data were misguided people saw Google and at the time, Yahoo was quite active with Hadoop and they could go and process all this historical data and try to find insights. So if you're just taking data and wandering around to see what you're going to find, I think you're working backwards. You want to look at the business first and figure out what problems are we trying to solve and then go find the data to support it. And then you can use all these wonderful tools, but much of it was kind of in the sandbox, playing around, drive your data opportunities with business needs first. And if you start from that end, you're more likely to have good solutions. The real time element is to change your thinking. You have to open up your mind. What if I could do something now instead of looking at consumer trends and figuring out how I can make an offer to this person next week, what if while they're in the moment, what if while I'm driving my Tesla to Tahoe, I'm completely connected. I'm an interconnect connection and I'm driving by various places. Why not have offers coming that at the next exit, Starbucks will give me a discount on my decaf mocha? Why not? Those are things that nobody would think of now, but you have to think about what if you could do it in the moment. So let me, let me one more, so let's add one more thing. What about offers now, getting off the next exit versus offers when you actually get to Tahoe and helping you trade off which ones you'd rather have? All part of the same problem. All possibilities, but offering them in real time. Chair, I want to use the last minute to get a different perspective from you because you sit on a lot of boards, you have a global view, you see a lot of things, you're involved in a lot of philanthropy. We touch briefly on it in Big Data SV. We're seeing and we're getting pulled in with theCUBE data philanthropy concepts, meaning people are being data rich in the sense that they either have data they've acquired or have access to data, that could be smart cities, could be other things, UN to solving hunger and poverty and other traditional philanthropy issues. Is there an enablement model? Can you share your view on this? Because we're seeing a collision course with business enablement and society enablement right now and people are trying to figure out what to do. Is it just typical philanthropy from a corporation, oh yeah, I'm donating some money, I've been involved in, yeah, yeah, chocolate output on the press release, so much more active in the moment like impact. Can you share what you're seeing? Because this is a trend that's emerging very quickly where it's not just so much philanthropy and donating, really involvement being present, being active with the data and active with the time. Over the last 20 years, in addition to all the boards I've been on, I've always done one board that I call double bottom line and it's actually not just philanthropy, it's actually double bottom line means for-profit company, not a nonprofit, for-profit, for social good. And the reason all of these that we've done have been double bottom line is we believe that if you build a for-profit company, it can have a bigger impact, it can have a more sustainable business. So just one very quick example, right now the company I'm most excited about that I'm working on in Nairobi, it's called Copia, you can look up at copiaglobal.com and what we're doing is we're bringing Amazon.com like services to the base of the economic pyramid. The two to four billion poorest people in the world would like to have just like you and me the ability to order online and have delivered to their home, only two small problems. They have no way to get online to order and nobody would ever deliver to their home. Piece of cake, right? So what we do is we go into the village, we find a business person, typically a woman, maybe has a small shop or something, we enable them with a tablet, we have a couple of thousand products, the villagers walk up, can order online, they pay with mobile money on their telephone, they're unbanked but they have a telephone and they have mobile money on Pesa. We get the order, we deliver 48 hours later just like Amazon Prime, they walk back, we deliver to the agent, they walk back, pick up their stuff and go home. Agent makes commission, they're typically doubling their income. The people in the village don't have to take the chicken bus and go into town, spend the whole day, they get city prices. We're having a big, big impact growing what will be a big company all through Africa and Asia, South America. So this is an opportunity to build a big, for-profit company with major social good, completely online, the data is the heart of it. We couldn't offer the service without an online system that went end to end. We're economically infeasible. And you see this trend continue, I see the economic impact and the quality of life with having goods and services is solid. And a lot of these places are just leapfrogging the U.S. They actually have more- Fine necessity, it's the mother of all invention, right? So technology is having an impact not just on all the businesses that we work with, but it's really impacting technology benefiting humanity worldwide. Jerry, thanks so much for spending the time to come back on theCUBE, really appreciate your insights. That's always fun. And your perspective, great to have you. Real thought leadership, but also experience and also real good grasp of the issues, thanks so much. Here to theCUBE live here in San Francisco, I'm John Furrier with Peter Burris with theCUBE. You're watching theCUBE here in San Francisco. It's always fun to come back to theCUBE because...