 Live from Miami, Florida, it's theCUBE. Covering IBM's data in AI forums. Brought to you by IBM. We're back in Miami and you're watching theCUBE's coverage of the IBM data and AI forum. Tony Hyam is here as a distinguished engineer for the digital and cloud business analytics at IBM. Tony, first of all, congratulations on being a distinguished engineer. That doesn't happen often. And thank you for coming on theCUBE. Thank you. So your area of focus is on the BI and the enterprise performance management space. And if I understand it correctly, a big mission of yours is to try to modernize those, make them self-service, make them cloud ready. How's that going? It's going really well. I mean, we use things like BI and enterprise performance management, but when you really boil it down, that's analysis of data and what do we do with the data that's useful that makes a difference in the world. And then there's planning and forecasting and budgeting, which everyone has to do, whether you are a single household or whether you're an Amazon or a Boeing, which are also some of our clients. So it's interesting that we're going from really enterprise use cases, democratizing it all the way down to single user on the cloud, credit card swipe, 70 bucks a month. So that was what used to work for Lotus, but Cognos is one of IBM's largest acquisitions in the software space ever. Steve Mills and his team architected a complete transformation of IBM's business and really got heavily into it. I think it was a $5 billion acquisition. Don't hold me to that, but massive one at the time and it's really paid dividends. Now, when all this sort of the 2010s came in, we said, oh, Hadoop's going to kill all the traditional BI, traditional EDW, that didn't happen. That these traditional platforms were a fundamental component of people's data strategies. So that created the imperative to modernize and made sure that there could be things like self-service and cloud ready, didn't it? Yeah, that's absolutely true. I mean, the workloads that we run are really sticky workloads, right? When you're doing your reporting, your consolidation, or your planning of your yearly cycle, your budget cycle on these technologies, you don't rip them out so easily. So yes, of course there's competitive disruption in the space and of course cloud creates an opportunity for workloads to be run cheaper without your own IT people. And of course, the era of digital software, I find it myself, I try it myself and I buy it without ever talking to a salesperson, creates a democratization process for these really powerful tools that's never been invented before in that space. Now, when I started in the business a long, long time ago, it was called the DSS Decision Support Systems. And they, at the time, they promised a 360 degree view of the business. That never really happened. You saw a whole new raft of players come in and then the whole BI and Enterprise Data Warehouse was going to deliver on that promise. That kind of didn't happen either. Sarbanes-Oxley brought a big wave of imperative around these systems because compliance became huge. So that was a real tailwind for it. Then Hadoop was going to solve all these problems. That really didn't happen. And now you've got AI. And it feels like the combination of those systems of record, the data warehouse systems, the traditional business intelligence systems, and all this new emerging tech together are actually going to be a game changer. I wonder if you could comment on that. Well, so they can be a game changer, but you're touching on a couple of subjects here that are connected, right? Number one is obviously the mass of data, right? Because data has accelerated at a phenomenal pace. And then you're talking about how do I then visualize or use that data in a useful manner and that really drives the use case for AI, right? Because AI in and of itself or augmented intelligence as we talk about is only useful almost when it's invisible to the user because the user needs to feel like it's doing something for them that's super intuitive. A bit like the sort of transition between the electric car and the normal car. That only really happens when the electric car can do what the normal car can do. And so with things like, imagine you bring a Hadoop cluster into a BI solution and you're looking at that data. Well, if I can correlate, for example, time, profit and cost, then I can create KPIs automatically and I can create visualizations. I know which ones you'd like to see from that or I can give you related ones and I can even automatically create dashboards because I've got the intelligence about the data and the knowledge to know how you might visualize it versus you have to manually construct everything. And AI is also going to, when you bring these disparate data sets together, isn't AI also going to give you an indication of the confidence level in those various data sets? So for example, your BI data set might be part of the general ledger or the income statement and be corporate fact, very high confidence level. Where sometimes you mentioned Hadoop, some of the unstructured data, maybe not as high a confidence level. How are customers dealing with that and applying that? First of all, is that a sort of accurate premise and how is that manifesting itself in terms of business? Yeah, so it is an accurate premise because in the world of data, there's the known knowns and the unknown knowns, right? The known knowns are what you know about your data. What's interesting about really good BI solutions and planning solutions, especially when they're brought together, right? Because planning and analysis naturally go hand in hand from the one user at 70 bucks a month to the enterprise client. So it's things like, what are your key drivers? So this is going to be the drivers that you know what drives your profit, but when you've got massive amounts of data and you've got AI around that, especially if it's AI that's got an ontology around your particular industry, it can start telling you about drivers that you don't know about. And that's really the next step is, tell me what are the drivers around things that I don't know. So when I'm exploring the data, I might see a key driver that I never even knew existed. So when I talk to customers, and I've been doing this for a while, one of the concerns they had, the criticisms they had of the traditional systems was, ah, the process is too hard. I got to go to, this has got a few guys I can go to, I got to line up, submit a request, by the time I get it back, I'm on to something else. I want self-serve beyond just reporting. How is AI and IBM changing that dynamic? Can you put these tools in the hands of users? Right, so this is about democratizing the cleverness. So if you're a big, broad organization, you can afford to hire a bunch of people to do that stuff. But if you're a startup or an SMB, then that's where the big market opportunity is for us. You know, abilities like, and we're building this into the software already today, is I bring a spreadsheet along. Spreadsheets, by definition, they're not rows and columns. Anyone could take a row and column spreadsheet and turn into a set of data, because it looks like a database. But when you've got different tabs and different sets of data, that may or may not be obviously relatable to each other, that AI ability to be able to introspect a spreadsheet and turn it into, from a planning point of view, cubes, dimensions and rules, which turn your spreadsheet now to a three-dimensional in-memory cube or a planning application. You know, our ability to go way, way further than you could ever do with that planning process, over thousands of people, is all possible now because we've taken all the hard work, all the lifty work out of it. So that three-dimensional in-memory cube, I like the sound of that. So there's a performance implication, obviously. And then there's, what else? Accessibility to more apps, more users, is that? Well, it's the ability to be able to process what if things on huge amounts of data. Imagine you're Boeing, right? How many parts does Boeing have? I don't know, three trillion, I'm just guessing, right? But if you've got three trillion, and you need to figure out, based on the latest hurricane report, how many parts you need to go ship to where that hurricane report is, you need to do a what if scenario on massive amounts of data in a second or two. So, you know, that capability requires an OLAP solution. However, the rest of the planet, other than OLAP people, bless them, who are very special people, don't know what OLAP is from a pop-tart. So democratizing it, right? To the person who says, I've got a set of data, and I still need to do what if analysis on things, and probably at large data, because even if you're a small company with massive amounts of data coming through, people click-stream me through your website, just for example, you know, what if analysis on putting a 5% discount on this product based on previous sales, how's that going to affect me from a future sale? So again, I think it's the democratizing as well as the ability to hit scale. So when you talk about cloud and analytics, how they've come together, what specifically IBM has done to sort of modernize it as platform, and I'm interested in what customers are saying, what's the adoption like? So, I manage the global cloud team, and we have nine on a thousand clients that are using cloud, the cloud implementations of our software. Growing actually, so actually more on two and a half thousand if you include the multi-tenant version. There's two steps in this process, right? When you've got an enterprise software solution, your clients have a certain expectation that your software runs on cloud just the way as it does on-premise, which means in practical terms, you have to build a single-tenant or managed cloud instance, and that's just the first step, right? Because getting clients to see the value of running the workload on cloud where they don't need people to install it, configure it, update it, troubleshoot it, and all that other sort of IT stuff that subtracts you from doing, running your business value, we do all that for you, but the future really is in multi-tenant and how we can get vast scale and also greatly lower cost. But the adoption's been great, clients love it. Can you share any kind of indication or is that all sort of confidential or what kind of metrics do you look at? So obviously we look at growth, we look at user adoption, and we look at how busy the servers are. I mean, let me give you, the best way I can give you is a number of servers, volume numbers, right? So we have 8,000 virtual machines running on SoftLayer or IBM Cloud for our clients. Business analytics is actually the largest client for IBM Cloud running those workloads for our clients. So it's, you know, the adoption's been really super hot and the growth continues. Interestingly enough, I'll give you another factoid. So we just launched last October, Cognos Analytics multi-tenant. So it is truly multi-tenant infrastructure. You try, you buy, you give your credit card, and away you go. And you would think, because we don't have software sellers out there selling it per se, that it might not adopt as much as people who are out there selling software. Well, in one year it's growing 10% month on month. So gradually it's 10% month on month. And we're at nearly 1,400 users now without huge amounts of effort on our part. So clearly there's market interest in running those softwares. And then they're not onesie twosies. These are six people per tenant. Some of people have 150 people per tenant on a multi-tenant software. So I believe that the future is dedicated as the first step to grow confidence that my on-premise investments will lift and shift to cloud, but multi-tenant will take us a lot further. So that's a proof point of existing customers saying, okay, I want to modernize. I'm buying in. They'll take a half step of the managed. Dedicated, yeah. And then obviously multi-tenant for scale and just way more cost efficient. Yes, very much so. All right, last question. Show us a little leg. What can you tell us about the roadmap? What gets you excited about the future? So I think the future, historically, planning analytics and cognitive analytics have been separate products, right? And when they came together under the BI logo in about a year ago, we've been spending a lot of our time bringing them together because, you know, you can fight in the BI space and you can fight in the planning space. And there's a lot of competitors here, not so many here, but when you bring the two things together, the connected value chain is where we're really going to win. But it's not only just doing it as a connected value chain, and it could be being being biased because I'm the former Lotus guy who believes in democratization of technology, right? But the market's showing us, when we create a piece of software that starts at 15 bucks for a single user, for the same power, mind you, right? Little, you know, less of the capabilities of 70 bucks for a single user for all of it, people buy it, so I'm in. Great, Tony, thanks so much for coming on theCUBE. It was great to have you. Brilliant, thank you. All right, and keep it right there, everybody. We'll be back with our next guest. You're watching theCUBE live from the IBM data and AI form in Miami. We'll be right back.