>> From Times Square, in the heart of New York City, it's theCUBE. Covering Imagine 2018. Brought to you by Automation Anywhere. >> Hey, welcome back everybody. Jeff Frick here with theCUBE. We're in Manhattan at the Automation Everywhere Imagine 2018. About 1,100 people talking about RPA, Robotics Process Automation, bots, really bringing automation to the crappy processes that none of us like to do in our day to day job. And, we're excited to have a practitioner. He's out in the field. He's talking to customers all the time. It's Weston Jones, and he's the global intelligent automation leader for EY. Weston, great to see you. >> Yeah, thank you, good to be here. >> Absolutely, so it's funny, you said you've been with these guys for a number of years, so when did you get started, how did you see the vision when nobody else saw it, and here we are five years later, I think, since you first met 'em. >> Oh, I know, it's just funny. I mean, years ago I saw Automation Anywhere at conferences. They were one of the small booths, just like everybody else was, talking about automation. I watched them for several years, and then I decided one year when we were looking at some of our offerings to bring in RPA and talk to our leadership about it, and kinda the light bulbs went off. So, from five, six years ago 'til today we've been working with them, and it's really amazing to see kind of how things have changed, and how the adoption has taken place. >> You know, it's such a big moment in a startup, especially software company, when you get a big global integrator like you guys to jump in, you know, advisory service. It's really hard to do. I've been in that position myself, and you guys don't make the move unless you really see a big opportunity. So, what did you see in terms of the big opportunity that made you, you know, basically bet your career on this vertical? >> Well, so when I went to our leadership, in the meeting I had our global shared services leader. So, we have 7,000 plus people on our shared services, and he was very skeptical. We had to do 20 plus proof of concepts with him, and HR, IT, finance, et cetera, to get him excited about it. Now, he's our biggest fan, and actually we promoted him to run our global internal automation team where now we think we're one of the largest users of automation. We're one of the biggest users within tax. We use Automation Anywhere within tax. We have over 750 bots working, and we have a goal to have 10,000 plus by 2022. So, we're really pushing the bar in scaling. >> From 750 to 10,000, what are we, 2018, in four years. >> In four years. That's our goal. >> So, where did you find the early successes, what kind of bots specifically, what type of processes are kind of right for people that are interested, see the potential, but aren't really sure kinda how to get started, or to get that early success? >> Yeah, I mean, it's just almost like anything else, the quick wins, you know. Start with things that are very rules-based, that have a lot of people, FTs associated with them. You know, our thing wasn't that we were actually eliminating FTs, we were just developing capacity, 'cause we're a company that's growing, so instead of hiring more and more people, we took all that mundane work out of people's jobs and allowed them to focus on things that were more value-added. So, the block and tackle stuff-- >> Like what? Like, give me a couple of, you know, just simple stuff-- >> well, we have like HR onboarding, you know, we onboard 60,000 people a year. HR onboarding is something that's very repetitive activity, logging in and out of multiple systems. And, it was something where we were hiring HR professionals that knew how to do talent management, that knew how to do all these things we really wanted them to do, but we had 'em focused on doing a lot of very transactional type activities. So, we said why don't we use the technology for that. Let's free these people up so they can then focus on developing talent, career ladders, other things that we really wanted them to focus on. Other things like, you know, payments, matching, and payment application, things like that, password resets, you know, a lot of stuff that you, I mean, you can just think of in your head. A lot of stuff in finance, a lot of stuff in HR and IT. Even our supply chain, too. We're doing like T and Es, we're doing a lot of automation in our T and E area. But, that to say, I mean, I've mentioned all back office things. We're also doing a lot of front office. So, for example, in our tax department we use almost exclusively Automation Anywhere to do tax returns for clients. And, we have, I think, over a million plus hours that we've eliminated using Automation Anywhere. >> Now, how do you Automation Anywhere a tax return? >> Well, tax return is a very complex set of rules, and you basically, once you kind of load the rules in for certain activities, it's stuff like pulling data from one system into another, you know, doing multiple taxed jurisdictions. >> Is it just like particular steps within that, you just kinda pick off one little process at a time, one little process at a time? >> True, and then you can also put in, you can do a nice interface in the front, and you can have people giving you the data, and then you let the automation then get the data to the right parts within the tax return. >> So, I'm curious in terms of the people that create the bots. Who are they, kinda what skill sets do they have, and do you see that changing over time as you try to go from 750, whatever it is, a 20x multiple, over four years? Do you see kinda the population of people that are able to create and implement the bots growing? How do you, kinda, managing the supply side on on that? >> We have a philosophy that 70% of it's process, 30% of it's technology. We're fortunate that in our advisory area across all the major functional areas, supply chain, HR, finance, et cetera, we have process experts. So, we use those process experts to get the process down, and then what we do is we have core development teams around the world. We have a big team in India, a big team in Costa Rica. We have a team in China, and elsewhere. And, those are the developers. And, so our process people map out the process and then hand that off to the developer. So, developers, you know, we basically, I mean, with Automation Anywhere's help, we've trained them to do the work and they've made it more and more, as time goes on, they made it easier and easier for them to develop bots. And, so We've been able to take people almost right out of college. We've hired some high school students. We take people that, you know, two thirds of the American population doesn't have a college degree, so we hire non-college degrees and teach them how to do this. Not that it's easy, and to be really good you have to have time and experience, but we can teach them to do these types of activities for us. >> That's amazing. So, I wonder if you can share what are some of the biggest surprises, you know, kind of implementation surprises, or ROI types of surprises that you found in implementing these 750. >> Yeah, so one thing I tell people about is if you talk about the Gartner Hype Curve, you go up and you fall into the valley of disillusionment, and, you know, there's gonna be four or five of those valleys that are gonna happen, and you just need to power through them because the technology is so compelling, and the benefits are so compelling. I mean, there's over a dozen benefits whether it's cost savings, improved security, better accuracy, whatever. So, some of the surprises were scaling. Like, when I talk about the DIPSS, the D-I-P-S-S, DIPPS, the first one is gonna be data. People are gonna realize that their data isn't quite there in order to do the more intelligent activities. The integration, so integrating the RPA with the more intelligent pieces of the IQ bot, and other things, how do you do those integrations, how do you take other tools outside of that and integrate them. The third is penetration. I mean, penetration is very small right now. What happens is people tend to look at a whole process that needs to be automated when what you need to do is you need to think about breaking those processes apart. Like FPNA, for example, may have a couple dozen steps to it, but there are pockets of steps that are very automatable. For example, pulling data, structuring it, normalizing it, getting it into some kind of report, that can all be done by automation, then hand it off to someone to do more cognitive activities. So, the penetration is very small right now, but will continue to grow. The savings, you know, have realistic expectations on savings. When this first came out of the door a lot of people were talking very, very high numbers. I mean, you can get it every once in a while, but, the saving numbers, just be realistic about that. And, the last part is scaling. We found scaling to be something that, you know, at the time when we were doing it, very few people had done it. So, to figure out how do you scale, and how do you develop a bot control room, how do you manage the bots, how do you manage the bots interfacing with people, how do you manage the bots interfacing with other technologies. It's a lot more to it than just putting the bot up and letting it work, because they need care and feeding ongoing, because it's not related to the Automation Anywhere technology, it's more of the other things it touches, like website changes, like upgrades to different systems that the bot has to execute with. Those are gonna constantly change and you just need to make sure you're adjusting the bot to actually work in those environments. So, those are kinda the four or five things that we've seen. And, when we go from 750, to 1,000, to 10,000, I mean, we think we're gonna see much more orchestration type things. You know, how do you orchestrate in a more automated way across the bots, the people, and then the other technology. >> Right, it's funny on the scale issue 'cause they were talking about, you know, how do you go from 10 bots, you got 750 to 10,000, and there's been a concept under it that they are a digital workforce, implying that you have to manage 'em like a workforce. You gotta hire 'em, you gotta train 'em, you gotta put 'em in place, you gotta kinda keep an eye on 'em, you gotta review 'em every now and then, and really it's an active management process, it's not just set and forget. >> Yeah, we're hoping that we'll have, I mean, we have some of this already, but we'll have bots managing bots. Well, bots auditing bots. We'll have bots orchestrating bots. That's all gonna eventually happen. I think we can do some of it today, but it's gonna be more and more common. The orchestration piece is really the thing that is gonna be new, that is gonna drive a lot of people this hard to scale. >> The other two consistent themes that you just touched on that we talked a little bit before we turned the cameras on, is Amara's Law, my favorite. You know, we overestimate the short term, which Gartner might call the Hype cycle, but we underestimate in the long term. Really, the other one is kinda just DevOps, and there's DevOps as a way to write code, but I think, more importantly, is DevOps as a culture, which is just look for little wins, little wins, little wins, little wins, little wins, and, before you know it, you've automated a lot and you're gonna start seeing massive returns on that effort versus the, oh, let's throw it in, we're gonna get this tremendous cost savings on day zero, day one, or day 10, or whatever it is. That's really not the strategy. >> Well, I think a lot of people maybe don't like to hear this, but it's a journey. I mean, you start out using the technology where you can. So, it's not a technology play, it's solving your biggest, most complicated problems, that's the key. And, whatever technology you need to do that, use that. So, you do the RPA, then you get more benefit when you add the IQ bots, and the intelligent stuff, and you get more benefit when you start adding, you know, technologies that are even ancillary, like Blockchain, IoT, and things like that. You'll get more and more kind of benefits from this technology. >> All right, Weston, well, thank you for sharing your stories. It's good to get it from the front lines. And, good luck on making 20,000 bots in four years. >> Thank you, thank you. >> He's Weston, I'm Jeff, you're watching theCUBE from Automation Anywhere Imagine 2018. Thanks for watching. (upbeat music)