>> Narrator: Live from Las Vegas it's The Cube covering Interconnect 2017. Brought to you by IBM. >> Okay welcome back everyone. We are live at the Mandalay Bay for IBM Interconnect 2017, The Cube's exclusive coverage. I'm John Frower, Dave Vellante, my co-host. Our next guest is Jim Casey and Michael Gilfix. Michael's the VP of process transformation and Jim is offering manager at IBM. Guys, welcome back to The Cube. >> Both: Thank you. >> So you guys had a big announcement on Monday, the digital assistant, so I've been craving a digital assistant since the little Microsoft little, you know, icon would pop up. >> Michael: You're talking about Clip, aren't you? >> The clip man. >> Don't talk about that. >> We don't like that. >> To me that was once called the digital assistant. It was a help button, but this is now, digital assistant is real automation, and you guys got a whole other take on this. It's totally cloud, cloud first. What's the digital assistant product that you announced? Take us through that. >> So here was our vision. What we found was in the modern, digital workplace, everyone is struggling to just keep up pace. Too many sources of information, and the information is buried everywhere. It's buried in emails, in spreadsheets, in documents. Many corporations have undertaken a BI project. In fact, there's an explosion of all these different dashboards that has all kinds of business data that they could go and see, so no one has the time to read all these things. Meanwhile, everyone in the modern world is trying to do 50 things at once and it's hard to figure out what is the best time to progress something and make progress? Our vision, so what we thought is wouldn't it be great if I could program this assistant, programmable by everyday business users, to watch for the things that matter to me and figure out when I should take action or take automated action on my behalf to save me time. >> So it's an interface, so it's software interface, cloud-based SAS, and the back end, does the user have to, what's the persona of the user that's using your product? >> Well, we want them to be used by non-developers, non-technical users, and so we thought really carefully about how you can teach your assistant these notions of skills, really point to tasks that can really make your life easier on a daily basis and they can pick anything that they like working with, that they can connect to, get the information from, and effectively assemble into these point-to tasks. >> Host: And the data sources are whatever I want them to be, explain how that works? >> Yeah, it can connect to common SAS applications. Those could be things like productivity suites, like G-Suite, they can be things like CRM systems, like Sales Force, campaign management systems like Marketo, and that's just in the beta that we just launched. And of course in the future, they'll be able to connect into their on-premise systems as well. >> So is it to replace the dashboards and all the wrangling that goes on? Most business users will have either a department that does all the data science or data prep for them, wrangling data sets, and then they get reports or spreadsheets or some BI dashboard. >> Yeah, we wanted the assistant to push the work to the user instead of the user having to go and spend time watching all these dashboards that really, they just didn't have time to do. And so the assistant takes all the heavy lifting of watching the data for you, figures out when action is needed, and then taps you on the shoulder. >> So Ginny Ramete was talking about that your customers want to own the data. So that's a great purpose, we buy into that mission, but a lot of the data is spread all over the place, so one of the problems that we're seeing in the big data world, now IOT complicates even further, is that data's everywhere, scattered, and the tools might have stacks and data wrangling within tools so you have complexity out there just on the scaffolding of how the data's managed. Is that part of the problem that you guys help solve? Because that seems to be a pain point. >> Yeah, and I think the amount of time that people spend just searching and aggregating and gathering information so they can figure out what to do, it's staggering. And when you think about the, it takes about two the three hours often for people to gather all the information that they need in order to make a real significant decision, every day, daily, you know operations. You're spending time in your email, you're building spreadsheets. Think of all the time you spend building a spreadsheet, wrangling data, you know. It's a productivity killer, and so a lot of the use cases that we look for, we'll ask our clients show me the ugliest spreadsheet that you use on a day-to-day basis for business operations. That's usually a starting point, or show me how many dashboards are you looking at and what are the decision you make off that? That's the stuff that we want to collapse into what the assistant can provide. >> So I got a use case for you, I'm a walking, I'm like everybody, right, so I've got my email, I've got five or six spreadsheets, Google Docs that I'm in every day all day, maybe there's a base camp, maybe there's a slack. I'm in Sales Force, all right, and then I got my social. >> Tool overdose. >> You just described the typical modern environment. All fragmented tools. >> And I'm in there and I'm like which browser is it, oh is it in Firefox, I'll put my Safari stuff I'll put over here, and I'll put my email in Mozilla, okay. It is just awful, it's a bloody nightmare, I get lost. I got to back up, hit the escape key, and go, okay, where am I, how do I find it again? >> Jim: It's connecting the dots. >> Okay, explain now how you can help me. >> So think of the things that you're looking for in all those different data sources. We're seeing the trend now. It's not about how can I just connect with things, it's how can I connect the dots? It's the actual business data inside of there, and how do I put that in a context that's relevant to you, what you're trying to do? You know, and a great example, we're working with one client who, they're moving, and a lot of people are doing this, they're moving from a point in time sale to being as a service, and in that kind of scenario, relationships with your clients really matter. And preventing customer churn is really important. So they have people who are responsible for making sure that people are not going to churn. That's a lot of dots to connect, right? So with the Digital Business Assistant, what we do is we look for those patterns that are really common that predict churn, but those things are scattered across your sales systems, your marketing systems, the website traffic, social media even, and we're able to combine all those things into a really consumable component called a skill. And then that individual person that's responsible for this set of customers can tailor it to their needs. So it's kind of like how you would buy a suit. When you go in and buy a suit, you don't get just the fabric laid out on a table and they cut it, right? You, most people don't anyway. (they laugh) >> I buy what's on the rack. I say "I want that one." >> Yeah, you walk in and you say that. >> I want what that is. >> 42 long, right? And they make a couple adjustments and then it's yours. >> All right, I'll take that suit up there, what's on the mannequin. >> They make a few adjustments and it's yours. Software should be the same way. You should be able to configure software in a few clicks. >> That's the whole thing, I mean, I joke about the mannequin but that's really kind of what hangs the perfect use case so that would be an automated example of an assistant model for you guys. Sometimes you just want everything to hang together for you, and sometimes you might want to go in and go look at the data. >> Yeah, and we see this across a lot of different industries, so things like customer service and sales and marketing, but we also see it in, let's say I'm a field technician, right? And I got to go out to an oil field. How do I know all the different patterns of information that might predict whether or not I need to, what I need to do when I'm out there. >> So you monitor my patterns, my behavior, and then ultimately train the model, or? >> Well you program it. You tell it what to watch for for you. So to give you an example of the kind of use case, to pick a specific use case, and we shared this again in sort of our unveiling on Monday. We shared the idea of a sales rep who is pursuing a given opportunity, and thinking about all the factors that went into their success and, you know, that sales rep has several different things they need to use to really maximize their chance of closing that deal. So one is they need to be responsive do their customer, and you know, like many different corporations out there who sell many different products and services, while you're busy working on the new opportunity, you've got to service the old. So when some issue comes up, you have to be responsive to it. Well, it's really hard while you're busy working on all these opportunities, to make sure that the issue's being resolved, that you're being responsive to your customer. Meanwhile, everybody in the corporation is coming up with new opportunities, new marketing brochures, new values in the product. And so is your rep knowledgeable about the latest and greatest products? So we imagine that you could teach your assistant how to watch some of this stuff for you and really help you to close your opportunity. And a very pointed example of the kinds of things that it should watch for you, I should be able to say something like hey, if I can have an active opportunity and then my customer goes and opens a service support ticket and that service support ticket hasn't been resolved in a week and meanwhile, I got a bunch of email coming from that client, of tone angry, notice the cognitive part there, about this particular product, and meanwhile I'm on the road and I'm not checking my email. Well, I have a catastrophe waiting to happen. So I can program my assistant to watch for these kinds of things. >> Does it do push notifications? >> Exactly, so you can then have it push to you, look, here's all the information about the active service thing, here's how long it was sitting there waiting for resolution, this is what's happened since, and you can immediately take action. >> So you're orchestrating basically signals that the user connects, like a Google alert on search is a trivial example, right? Someone types, a result comes on Google, you get an email. Here, you're kind of doing that-- >> But it's proactive. You tell your assistant to proactively watch it for you, and that's a unique technology that we developed in-house. Because it's watching all these events happening in the enterprise and figuring out when that thing becomes actionable. >> And the user would know where to look, because like Dave's spreadsheet might say "hey, cash balance" or you know, sales trend, this rep and then something happens, and he can get that pushed to him from three different disparate side-load apps, that's pretty much what it is. >> That's right. >> Okay, so give us the status on the beta right now. It's a beta, so it's sign-up required. Okay, and the requirements to implement it, if you get through the beta, is just log in to a portal? It's a SAS model and then do the connectors? >> So the first thing you do, you go to IBM.com/assistant. You can sign up to. >> That, by the way, might be the easiest URL I think we ever came up with. I'm pretty sure that one's going to be memorable. >> Yeah, so you just go to that site, you sign up, you give us a little bit of information, your email, how to contact you and we'll put you on the waiting list, and what we're going to be doing is opening up more seats as we go through over the next couple weeks, and then we plan in the near term here to make it available as an open beta that you could see, and you'll see that inside of Bloomix as a tile inside of Bloomix. >> And here's the thing, we're doing something really different in the marketplace. This is a very different kind of offering, really targeting, again, non-technical people, this proactive situational awareness that your assistant can do, uses your data, built-in intelligence, intelligence that can customize to the way you work, guide you to the next best action. We have an incredible vision for this. The idea behind the beta is to start getting feedback. We worked very closely with early customers in the initial design and development. We want to open that up and get even more feedback and ideas on this kind of technology. >> So how is this different from Watson's discovery services that they have? I can imagine that you're building on Watson. Is it the cognitive piece within IBM, or is this kind of, I mean how would a customer figure that out, or just more of a-- >> Yeah, so I can give you an example. So we have one of our prototypes that we're actually taking some of the components of Watson discovery service and we package that up as a skill inside of your assistant, and it's a specific implementation, so what it allows you to do in this case is it'll look at your email and it'll look for specific entities, like a customer that matters to you, and if I get three emails of negative sentiment from a customer where I also have an open opportunity in the last week, that's a pattern I want to know about, right? Or we can start to correlate with all sorts of different things, so I think what you're going to see is these skills that we make available with the digital business assistant really up, take consumability of these really, really powerful technologies around cognitive and cloud. We take that to the next level. >> That's the key, how do we make Watson tailorable and put in the hands of every knowledge worker in every company? >> Host: So I presume you guys are dog fooding this personally, is that right? >> We have plans to do that, yes. >> Host: Oh, you haven't started yet? >> Sampling our own champagne. >> But we are, yes. >> He always gets called on that. >> We will be using it, yes. >> We created that champagne. >> We're beer drinkers, that's it, beer. >> We're going back to dog food, we eat beer, we should drink our own beer now. We created that with all our boost men, remember? (laughs) >> So get back to the status of the product. So it's got some Watson capability, but this is for the user to use. I don't have to get IT involved? >> Jim: That's right. >> This is where the user takes a personal productivity approach, and you bring in some Watson-- >> A user may not even know that they're using some of these Watson capabilities. To the end user, what do you want it to do for me? Well, I want it to tell me if, uh, if I think a customer might be upset with me. Well, that might be a combination of a lot of different things, but it just makes it really consumable and easy for people. >> So where do you guys sit within IBM? Because now there's like, because this is a really cool user tool, so is this part of Watson? >> Jim: We think so. >> Is it part of the Watson team? >> Well, honestly our organization doesn't really matter, I mean, we're working with teams across IBM as a whole. It's a great opportunity to take this technology and really reach a whole set of new use cases, I think, across the company, and we want to integrate Watson technology to, like we were saying, really make it easy for the end-user to go and access it. >> Any plans around developer outreach? >> Well, we will, I think, later this year, one of the things we envisioned really early on is that people are going to want to have pre-built skill sets, and that's a great opportunity to build an incredibly powerful ecosystem and we've been in discussion with a lot of our partners about how to do that. >> Well you guys are API based, so this is a beautiful thing, right? >> Well we're going to start to open up some SDKs to our partners, to others, and that's going to allow them to extend the assistant and really create even more powerful industry content. >> You know, the business model of reducing the steps it takes to do something and saving people time, making it easy to use is a magical formula of success. >> And not even just less steps, it's less time reading things, less time sifting through information so you can spend time on stuff that matters. >> Just email by itself, I mean, Dave, your example was the best, because I know, we live that. But we have a multitude of tools and sometimes it just organically goes, because the one guy like, you know, this tool set, or now I got-- >> So do you want to do the deal now or? >> Right, that's what I'm saying, they should be signing up. >> So do we get paid? (they laugh) >> We're already both signed up. We have a testimonial. >> If you can't get it, how can we get it? >> We'll kick the tires on it, and uh, but the thing that gets my excitement is potential for API integration. Because if I know I can the automation to a whole other level and the use cases start to patternize in the enterprise, then it can get interesting. All right guys, thanks so much. What's going on here with the show, what else is happening for you guys? Share some stories for the folks that aren't here, that are watching on IBM Go right now. What's the vibe at the show this week? >> Well, it's been a great vibe. We've had a chance to share some incredible success stories, so in addition to the unveiling of this particular product, on Monday we had a chance for one of our marquee clients to share their story, and I'll tell you a little bit about what they did. It was at the National Health Service of the UK. Part of their blood and transplant, and we were fortunate enough to have Aaron Powell, who's the chief digital officer there, share their story of using process technology to improve the speed at which they get organs in the hands of recipients, and they did it on the cloud. And the results they obtained were unbelievable. So the before and after, they had staff at 2am, writing lists of high-risk patients and how to map their donors and he kidded us not, that when someone's priority changes, they would wipe the board and reset things. And these are people's lives that are at stake in the matching process. >> And they're tired, human error is huge. >> Human error, absolutely, and by the way, when you look at the end-to-end process, there was something like 90 steps if I remember, 96 steps I think end-to-end. All of which were very manual and error-prone, and error-prone means risk. And they were able to improve organ allocation by 3x, so 3x faster, they automated something like 58% of the steps, reducing propensity for manual error, and what he shared in his story is, they successfully a few months ago did the first heart transplant on the cloud. >> Host: Wow, that's amazing. >> So it's an amazing, amazing story. >> That's a great story, yeah. Did he say that in the session? >> He did, actually, he said that. >> That's actually a good thing to chase down for a great blog post, that would be phenomenal. It would have been covered yet on the news? >> So we're going to post actually the video of it online so people can also see him live presenting his story, it was unbelievable. >> Make sure you send me the link. The other thing that they could apply there is two-block chain, I mean some of the block chain stuff coming out is going to be really interesting. >> Absolutely, and we're working very closely with that team to really leverage this kind of process technology, take people's business operations and connect that in to this feature network that's going to power businesses. >> CRM is the human supply chain, I mean, but now extend it out to the internet of things. I mean, it's interesting how this could play out. Guys, thanks so much for coming on The Cube. Thanks for sharing the insight, congratulations on the launch. I just signed up for the beta while we were talking. >> Dave: Me too, so let us cut the line. >> Done. >> We need it. Perfect use case, we need help. It's The Cube, of course, no help here, great guests here on The Cube. I'm John Frower, Dave Vellante, more great coverage, stay with us. Day three of Interconnect 2017, we'll be right back. (techno music)