>> Hi, I'm Peter Burris, welcome to this Cube Conversation, where we bring some of the best ideas in the industry to the Wikibon, SiliconANGLE communities as a way of catalyzing further conversation about some of the changes, and some of the opportunities that are presented by tech in a world that's digitally transforming. We're being joined today by Bobby Patrick, who's the CMO of a company called UiPath. UiPath is one of the leaders in a technology known as robotic process automation. And we're going to talk about the problems, the solutions, and the directions forward with this, what we regard as a very, very important technology on the rise. Bobby, welcome to the Cube. >> Hi Peter, great to be here. >> So Bobby, let's start with who are you? Tell us a little bit about yourself. >> Sure, Bobby Patrick, CMO, UiPath. I was CMO of the cloud business HPE prior, a Cube alumni, I guess we call them, right? And had a history of startups, in SQL space, and open source, and different transformational companies. But I was really intrigued by the idea of software robots over the last six months, and joined UiPath, and it's just been an amazing adventure to see this kind of amazing technology really deliver outcomes for companies faster than I've ever seen happen in tech in my history. >> Well let's talk about that, so, every technology that's going to be successful has to target a problem, a group of people who are trying to solve the problem, and a set of returns. So let's talk about, what is the problem that UiPath and related technologies are actually trying to address? >> Let me start at the macro level first of all. Robotic process automation is instructive trend for digital transformation. For the last three, five, seven years, vendors and companies have said, "I want to go digital." Digital, digital, digital. Right? And going digital's really difficult. Back office processes and front office processes are complex, there's a lot of human interaction involvement in it, and in the past, the way to tackle that would be IT would engage in a significant integration project, an IT project, would purchase equipment, technology, and try to rebuild and redefine the process. Robotic process automation tackles that problem of digital transformation in a very different way. It says forget the idea of going and redoing all of the systems in the data. Let's just take it from the human perspective. Let's see how does a human engage in the process today? And let's completely mimic that human interaction. Let's do that in a way that has complete accuracy, actually higher compliance, and do that in a way that's very simple. And so our technology, which has been built over a number of years, has now perfected the ability to actually replicate and emulate a human user, interacting with multiple systems, something they do every day, in the enter process, move data around, analyze data, look at the data for context, and then execute that process continuously. And that has resulted in an industry that is booming. Forrester said last year it was a $250 million industry, it'll be a $3 billion industry in three years. And UiPath is the fastest growing company in what's called RPA. >> So let me break that down a little bit. So a couple of general principles. The computing industry has always been focused on how we can substitute technology, specifically in the form of programs, for labor. And when we do that, we're able to reduce errors, we're able to speed up the activities, we have derivative opportunities to integrate things we never did before. So let me see if I got this right. What you just described is, instead of trying to substitute for that labor, by creating a net new program that has a whole bunch of static elements, we're actually going to turn these tools and apply them to the question of, how do people do things, let's substitute for the things that people do, start there. Have I got that right? >> Yeah, I mean imagine a contact center in a customer service operation for an airline. They receive an email from a customer that's complaining, about a bad flight, that contact center specialist has to look at a variety of systems to determine your status, what did we give you recently, in two days they will respond. Maybe the person's name is Michelle, and they'll respond in two days, and give you something. Well now, Michelle the software robot can do that exact same thing in a matter of minutes, with complete accuracy, and you can apply machine learning to it, and AI to it, to determine, 95% of the time, if I provide this in this situation, I'll have a happy flier, a happy member. That's what we're talking about here, which is software robots that essentially are perfecting these complex business processes both in the back office and in the front office. Often where lots of documentation's involved, lots of different systems are involved, and a human has to interact between all of those over and over. >> So, RPA, effectively, robotic process automation, effectively provides a means to mimic the work of a person. Because that's where most process engineering is done. Mimic the work of a person, codify it in a way that actually leads to a better business outcome. So that's what it is. Now take us through how does this work? Who are the people involved in the process, what does the technology do for them, how long does it take? Give us a little bit about it. >> So one of the beauties to RPA is that this doesn't require deep engineering talent to be able to build a software robot, and execute it. In fact, some of the breakthroughs in the technology that's been created at UiPath are a studio designer, which looks much like Visio, where you can drag and drop a workflow. Subject matter experts are becoming the next coders, really. Where they can actually design the workflow. There's actually a recorder function that can actually record the user-- >> That's where it starts. It starts with the recorder looking at how people are doing things and picking up that, and turning that in semantics that are meaningful to-- >> Yeah, defining a workflow. Which has exceptions, and handling. I should mention, when you're creating, or you're automating a process, there are really two kinds of robots you're designing. One is one called an attended robot, where along that workflow that robot's going to stop and ask a human a question. It's about a third of the market right now. So the process executes and a human might have to fill in some gaps alone the way. An unattended robot can run in the cloud, or on a VM in an on-premises data center, and execute that process behind the scenes, over and over and over.. So you kind of, you're building one of two. And UiPath supports both attended and unattended robots. But yes, you're designing the workflow, that workflow may interact with a variety of systems, you might receive an email and read the email, the robot reads the email. You might actually log into salesforce.com to find out if they're a customer. You might actually upload the email as a artifact, you might then download some information, put it into a PDF, and send it on. That's an example of a workflow that you're automating end to end. >> So we've got that workflow designed, what are we doing next with it? And who's doing it? >> So this is one of the beauties, too. One of the challenges in IT is projects take a long time. But in RPA-- >> Peter: And they fail. >> And they fail. What's interesting in RPA, what we've proven now, is you can pretty much begin to digitize a process in a matter of weeks. The outcomes are almost immediate. And payback periods are often six months or less. An RPA project almost self-funds itself, which is one reason why I think this is taking off so fast as well. >> So if we want to get a payback period in six months, it means that the whole notion of how fast does it take to get a group up and running on this, becomes crucial. So what is RPA typically, who's it typically targeting? Is it a professional software developer, someone with no technology expertise, business analysts, where-- >> Business analysts, definitely. You're talking line of business. You're talking finance operation. There's a lot of innovation in finance operations. How do I improve my ability to handle invoice reconciliation, and manage purchase orders, and all that paperwork and movement of data. >> So these are people that are familiar with workflows, they're familiar with process design, et cetera, but may not be familiar with coding >> These are subject matter experts. They're not familiar with coding, but they know the process really well. They know what to do when there's an exception, they know what to do in a sequence of events, and so that's why we often say subject matter experts are the next big coders, because they can actually go and learn, in fact UiPath, we have a program called an academy. Academy is on our site, we launched it last April. 35,000 people have been certified already. These are typically business analysts who go get trained, online, self-led, get a diploma, a foundational or an advanced diploma, and are RPA developers, in fact. Now you can go deep. There's C#, you can develop and you can go deep behind it. So I'm not saying there's not some ability to go really deep in certain development. But for the most part, you're a finance operation, you're an HR operation. I'll give you an example of one that just popped up yesterday. A company called West Monroe, they're a consulting firm in Chicago, they announced that they built Rosie the Robot. And Rosie the Robot, with UiPath, is a robot that onboards all their new employees. And they're doing a lot of M and A. They're growing really fast. And onboarding all the employees was a task that required a lot of people to do a very system data intensive process. Now Rosie does all that for them. Very simple example, you can then kind of zoom out and realize that really every process, most processes, have some kind of human interaction repetitiveness to them which a robot can either assist a knowledge worker, or can actually execute that entire process flawlessly. >> Now, we're not talking about technology that's really esoteric, that requires an enormous net new experience in learning from an operations staff. We're talking about technologies that can be targeted specifically to a problem, and end up generating artifacts that are familiar to what's currently in place. Have I got that right? >> Yeah, and I think what's important, so enterprise RPA really addresses two sides. One, the business side, that's trying to digitize a process and automate, maybe for cost savings, but more important than trying to apply AI. And get smarter, right? And so that's the business side. It also addresses the IT side, which is, okay, I've got to secure these robots, how do they scale, how do I manage and govern them, imagine having thousands and thousands of robots. I'll give you another example. NASA. The first robots that NASA launched, the first one, they named George Washington. And George did a-- >> Good name. >> Bunch of work for the finance group, and they got really comfortable with George. They'd walk in in the morning and say, "What has George done for me last night?" Which is awesome. But George was onboarded just like a human worker. Meaning he has to log in to different systems, just like a human worker, and by the way, his password expires every 90 days. So how do they solve that? They created the boss robot. And the boss robot's name? Constitution. Constitution changes George's password every 90 days. That's what's happening here, is you're building out your digital workforce. So IT worries, then, about how do I manage and secure, and scale, so we think about scalability, and big scale is a big challenge, but opportunity that we're focused on, tens of thousands of robots that companies will have. We often say it'll ultimately be one robot per every employee. >> So we have not, or I think you've mentioned the word, or the phrase, "AI" just once. So this is utilizing similar kinds of concept. The attended versus unattended, for example. How you go about training. But I got to believe that there's going to be a roadmap for integrating a whole bunch of these new technologies, that are capable of providing even more degrees of freedom, more functionality, how is RPA and some of these new technologies going to intersect over the course of the next few years? >> This is a really, really important question. So RPA, and enterprise RPA and UiPath, we believe it's a platform. So once you digitize that process, you can then do things with it. We've opened APIs, it's very extensible, you can integrate with a conversation API of Watson, integrate with a chat bot and have the robots do the back end work. At Exxon, they're doing IOT, and deploying sensors left and right, but all the systems in the back end are legacy systems and Excel spreadsheets. So the robots actually are the back end, supporting the deployment of IOT on the front end. So you have this amazing combination. But what people really want to do, then, is they want to look at that process and say, "How do I make it smarter? "How do I improve the productivity over time?" It's great to get that initial bump of perhaps cost savings, when you think about the robot doing, what eight FTEs did, the one robot does. So that's one thing. But the bigger thing is being able to apply data science to the process, looking for ways to mine the process, to think about how can I do the execution better? And that's when we apply machine learning to a process. Where we can actually look, instead of having a rule, in the process that executes, you actually have the experience. Where you say, oh 90% of the time it happens this way. So I'll fill the field in, instead of going and tracking down an empty field. So you can really get smarter, and really improve productivity. Peter, this is all about productivity. GE's a great example of one, that spoke at a conference of ours recently. And the first nine months, they had $25 million of productivity from the robots. The next nine months, $150 million. But this is not about cost-cutting or employees. This is about, actually, hiring. This is about getting smarter in every process. This is about eliminating errors completely. >> Well, productivity is not just a function of the denominator, which is cost. It's also a function of the work that you can perform. And so what you're saying is that utilizing these technologies, while it may displace certain laborious tasks, nonetheless, it's automating and improving the quality of a whole bunch of others, which allows people to go off and do net new things that perhaps are better in service to customers, for example. >> Yeah, one of the fast savings we're seeing from our customers is that they're actually able to use robots to fill the gap of being able to hire new employees. So that, in Japan, here's a great, Japan's almost a unique market. Japan RPA, and UiPath is used on some of the world's largest RPA projects, like SMBC Bank, or Dentsu, the advertising agency, company, there, they're using robots to address two things. One, the decaying population, so robots are filling the gap. And also, two, suicide. There are very high suicide rate because they work these amazing hours every week. Well they're actually using robots to reduce the number of hours, as the robots complement the work of the workforce in Japan. So we're really seeing, interestingly enough, is that robots are actually filling the gaps, and beginning to do work of a workforce that maybe you wish you could hire but you can't hire. So I think that trend is what we're going to see more of in 2018. >> Excellent. So, Bobby, thank you very much for coming on the Cube, here in our Palo Alto studios, and talking to us about RPA, robotic process automation. We heard a little bit about what is it, how does it work, what's the impacts of using it, and obviously, UiPath and yourself as a increasing or emerging force within an important new marketplace for enterprises and users who are trying to increase their productivity. >> Thank you, Peter. >> Once again, this is Peter Burris, in a Cube Conversation with thought leader Bobby Patrick at UiPath. Bobby, again, thanks for coming. >> Bobby: Thank you.