(upbeat music) >> Announcer: theCUBE presents UiPath Forward5. Brought to you by UiPath. >> Hi everybody, we're back in Las Vegas. We're live with Cube's coverage of Forward 5 2022. Dave Vellante with Dave Nicholson Ted Kumer this year is the Executive Vice President, product and engineering at UiPath. Brought on to do a lot of the integration and bring on new capabilities for the platform and we've seen that over the last several years. And he's joined by Dr. Edward Challis, who's the co-founder of the recent acquisition that UiPath made, company called Re:infer. We're going to learn about those guys. Gents, welcome to theCUBE. Ted, good to see you again. Ed, welcome. >> Good to be here. >> First time. >> Thank you. >> Yeah, great to be here with you. >> Yeah, so we have seen, as I said, this platform expanding. I think you used the term business automation platform. It's kind of a new term you guys introduced at the conference. Where'd that come from? What is that? What are the characteristics that are salient to the platform? >> Well, I see the, the evolution of our platform in three chapters. You understand the first chapter, we call that the RPA chapter. And that's where we saw the power of UI automation applied to the old problems of how do I integrate apps? How do I automate processes? That was chapter one. You know, chapter two gets us to Forward3 in 2019, and the definition of this end-to-end automation platform you know, with the capabilities from discover to measure, and building out that core platform. And as the platform's progressed, what we've seen happen with our customers is the use of it goes from being very heavy in automating the repetitive and routine to being more balanced, to now where they're implementing new brought business process, new capability for their organization. So that's where the name, Business Automation Platform, came from. Reflecting now that it's got this central role, as a strategic tool, sitting between their application landscape, their processes, their people, helping that move forward at the rate that it needs to. >> And process mining and task mining, that was sort of the enabler of chapter two, is that right? >> Well, I'd say chapter two was, you know, first the robots got bigger in terms of what they could cover and do. API integration, long running workflows, AI and ML skills integrated document processing, citizen development in addition to professional development, engaging end users with things like user interfaces built with UiPath apps. And then the discovery. >> So, more robustness of the? Yeah, okay. >> Yeah. Just an expansion of the whole surface area which opened up a lot of things for our customers to do. That went much broader than where core RPA started. And so, and the other thing about this progression to the business automation platform is, you know, we see customers now talking more about outcomes. Early on they talk a lot about hours saved and that's great, but then what about the business outcomes it's enabling? The transformations in their business. And the other thing we're doing in the platform is thinking about, well, where can we land with solutions capabilities that more directly land on business, measurable business outcomes? And so we had started, for example, offering an email automation solution, big business problem for a lot of our customers last year. And we'd started encountering this company Re:infer as we were working with customers. And then, and we encountered Re:infer being used with our platform together. And we saw we can accelerate this. And what that is giving us now is a solution now that aligns with a very defined business outcome. And this way, you know, we can help you process communications and do it efficiently and provide better service for your customers. And that's beginning of another important progression for us in our platform. >> So that's a nice segue, Ed. Tell about Re:infer. Why did you start the company? >> Right, yeah, so my whole career has been in machine learning and AI and I finished my PhD around 2013, it was a very exciting time in AI. And me and my co-founders come from UCL, this university in London, and Deep Mind, this company which Google acquired a few years later, came from our same university. So very exciting time amongst the people that really knew about machine learning and AI. And everyone was thinking, you know, how do we, these are just really big breakthroughs. And you could just see there was going to be a whole bunch of subsequent breakthroughs and we thought NLP would be the next breakthrough. So we were really focused on machine reading problems. And, but we also knew as people that had like built machine learning production systems. 'Cause I'd also worked in industry that built that journey from having a hypothesis that machine learning can solve a problem to getting machine learning into production. That journey is of painful, painful journey and that, you know, you can see that you've got these advances, but getting into broad is just way too hard. >> So where do you fit in the platform? >> Yeah, so I think when you look in the enterprise just so many processes start with a message start with a no, start with a case ticket or, you know, some other kind of request from a colleague or a customer. And so it's super exciting to be able to, you know, take automation one step higher in that process chain. So, you could automatically read that request, interpret it, get all the structured data you need to drive that process forward. So it's about bringing automation into these human channels. >> So I want to give the audience a sense here. So we do a lot of events at the Venetian Conference Center, and it's usually very booth heavy, you know, brands and big giant booths. And here the booths are all very small. They're like kiosks, and they're all pretty much the same size. So it's not like one vendor trying to compete with the other. And there are all these elements, you know I feel like there's clouds and there's, you know, of course orange is the color here. And one of the spots is, it has this really kind of cool sitting area around customer stories. And I was in there last night reading about Deutsche Bank. Deutsche Bank was also up on stage. Deutsche Bank, you guys were talking about a Re:infer. So share with our audience what Deutsche Bank are doing with UiPath and Re:infer. >> Yeah, so I mean, you know, before we automate something, we often like to do what we call communications mining. Which is really understanding what all of these messages are about that might be hitting a part of the business. And at Deutsche Bank and in many, you know, like many large financial services businesses, huge volumes of messages coming in from the clients. We analyze those, interpret the high volume query types and then it's about automating against those to free up capacity. Which ultimately means you can provide faster, higher quality service because you've got more time to do it. And you're not dealing with all of those mundane tasks. So it's that whole journey of mining to automation of the coms that come into the corporate bank. >> So how do I invoke the service? So is it mother module or what's the customer onboarding experience like? >> So, I think the first thing that we do is we generate some understanding of actually the communications data they want to observe, right? And we call it mining, but you know, what we're trying to understand is like what are these communications about? What's the intent? What are they trying to accomplish? Tone can be interesting, like what's the sentiment of this customer? And once you understand that, you essentially then understand categories of conversations you're having and then you apply automations to that. And so then essentially those individual automations can be pointed to sets of emails for them to automate the processing of. And so what we've seen is customers go from things they're handling a hundred percent manual to now 95% of them are handled basically with completely automated processing. The other thing I think is super interesting here and why communications mining and automation are so powerful together is communications about your business can be very, very dynamic. So like, new conversations can emerge, something happens right in your business, you have an outage, whatever, and the automation platform, being a very rapid development platform, can help you adapt quickly to that in an automated way. Which is another reason why this is such a powerful thing to put the two things together. >> So, you can build that event into the automation very quickly you're saying? >> Speaker 1: Yeah. >> Speaker 2: That's totally right. >> Cool. >> So Ed, on the subject of natural language processing and machine learning versus machine teaching. If I text my wife and ask her would you like to go to an Italian restaurant tonight? And she replies, fine. Okay, how smart is your machine? And, of course, context usually literally denotes things within the text, and a short response like that's very difficult to do this. But how do you go through this process? Let's say you're implementing this for a given customer. And we were just talking about, you know, the specific customer requirements that they might have. What does that process look like? Do you have an auditor that goes through? And I mean do you get like 20% accuracy, and then you do a pass, and now you're at 80% accuracy, and you do a pass? What does that look? >> Yeah, so I mean, you know when I was talking about the pain of getting a machine learning model into production one of the principle drivers of that is this process of training the machine learning model. And so what we use is a technique called active learning which is effectively where the AI and ML model queries the user to say, teach me about this data point, teach me about this sentence. And that's a dynamic iterative process. And by doing it in that way you make that training process much, much faster. But critically that means that the user has, when you train the model the user defines how you want to encode that interpretation. So when you were training it you would say fine from my wife is not good, right? >> Sure, so it might be fine, do you have a better suggestion? >> Yeah, but that's actually a very serious point because one of the things we do is track the quality of service. Our customers use us to attract the quality of service they deliver to their clients. And in many industries people don't use flowery language, like, thank you so much, or you know, I'm upset with you, you know. What they might say is fine, and you know, the person that manages that client, that is not good, right? Or they might say I'd like to remind you that we've been late the last three times, you know. >> This is urgent. >> Yeah, you know, so it's important that the client, our client, the user of Re:infer, can encode what their notions of good and bad are. >> Sorry, quick follow up on that. Differences between British English and American English. In the U.K., if you're thinking about becoming an elected politician, you stand for office, right? Here in the U.S., you run for office. That's just the beginning of the vagaries and differences. >> Yeah, well, I've now got a lot more American colleagues and I realize my English phrasing often goes amiss. So I'm really aware of the problem. We have customers that have contact centers, some of them are in the U.K., some of them are in America, and they see big differences in the way that the customers get treated based on where the customer is based. So we've actually done analysis in Re:infer to look at how agents and customers interact and how you should route customers to the contact centers to be culturally matched. Because sometimes there can be a little bit of friction just for that cultural mapping. >> Ted, what's the what's the general philosophy when you make an acquisition like this and you bring in new features? Do you just wake up one day and all of a sudden there's this new capability? Is it a separate sort of for pay module? Does it depend? >> I think it depends. You know, in this case we were really led here by customers. We saw a very high value opportunity and the beginnings of a strategy and really being able to mine all forms of communication and drive automated processing of all forms of communication. And in this case we found a fantastic team and a fantastic piece of software that we can move very quickly to get in the hands of our customer's via UiPath. We're in private preview now, we're going to be GA in the cloud right after the first of the year and it's going to continue forward from there. But it's definitely not one size fits all. Every single one of 'em is different and it's important to approach 'em that way. >> Right, right. So some announcements, StudioWeb was one that I think you could. So I think it came out today. Can't remember what was today. I think we talked about it yesterday on the keynotes anyway. Why is that important? What is it all about? >> Well we talked, you know, at a very top level. I think every development platform thinks about two things for developers. They think, how do I make it more expressive so you can do other things, richer scenarios. And how do I make it simpler? 'Cause fast is always better, and lower learning curves is always better, and those sorts of things. So, Re:infer's a great example of look the runtime is becoming more and more expressive and now you can buy in communications state as part of your automation, which is super cool. And then, you know StudioWeb is about kind of that second point and Studios and Studio X are already low code visual, but they're desktop. And part of our strategy here is to elevate all of that experience into the web. Now we didn't elevate all of studio there, it's a subset. It is API integration and web based application automation, Which is a great foundation for a lot of apps. It's a complete reimagining of the studio user interface and most importantly it's our first cross-platform developer strategy. And so that's been another piece of our strategy, is to say to the customers we want to be everywhere you need us to be. We did cross-platform deployment with the automation suite. We got cross-platform robots with linear robots, serverless robots, Mac support and now we got a cross-platform devs story. So we're starting out with a subset of capabilities maybe oriented toward what you would associate with citizen scenarios. But you're going to see more roadmap, bringing more and more of that. But it's pretty exciting for us. We've been working on this thing for a couple years now and like this is a huge milestone for the team to get to this, this point. >> I think my first conversation on theCUBE with a customer was six years ago maybe at one of the earlier Forwards, I think Forward2. And the pattern that I saw was basically people taking existing processes and making them better, you know taking the mundane away. I remember asking customers, yeah, aren't you kind of paving the cow path? Aren't there sort of new things that you can do, new process? And they're like, yeah, that's sort of the next wave. So what are you seeing in terms of automating existing processes versus new processes? I would see Re:infer is going to open up a whole new vector of new processes. How should we think about that? >> Yeah, I think, you know, I mean in some ways RPA has this reputation because there's so much value that's been provided in the automating of the repetitive and routine. But I'd say in my whole time, I've been at the company now for two and a half years, I've seen lots of new novel stuff stood up. I mean just in Covid we saw the platform being used in PPP loan processing. We saw it in new clinical workflows for COVID testing. We see it and we've just seen more and more progression and it's been exciting that the conference, to see customers now talking about things they built with UiPath apps. So app experiences they've been delivering, you know. I talked about one in healthcare yesterday and basically how they've improved their patient intake processing and that sort of thing. And I think this is just the front end. I truly believe that we are seeing the convergence happen and it's happening already of categories we've talked about separately, iPass, BPM, low-code, RPA. It's happening and it's good for customers 'cause they want one thing to cover more stuff and you know, I think it just creates more opportunity for developers to do more things. >> Your background at Microsoft probably well prepared you for a company that you know, was born on-prem and then went all in on the cloud and had, you know, multiple code bases to deal with. UiPath has gone through a similar transformation and we talked to Daniel last night about this and you're now cloud first. So how is that going just in terms of managing multiple code bases? >> Well it's actually not multiple Code bases. >> Oh, it's the same one, Right, deployment models I should say. >> Is the first thing, Yeah, the deployment models. Another thing we did along the way was basically replatform at an infrastructure level. So we now can deploy into a Kubernetes Docker world, what you'd call the cloud native platform. And that allows us to have much more of a shared infrastructure layer as we look to deliver to the automation cloud. The same workload to the automation cloud that we now deliver in the automation suite for deployment on-prem or deploying a public cloud for a customer to manage. Interesting and enough, that's how Re:infer was built, which is it was built also in the cloud native platform. So it's going to be pretty easy. Well, pretty easy, there's some work to do, but it's going to be pretty easy for us to then bring that into the platform 'cause they're already working on that same platform and provide those same services both on premises and in the cloud without having your developers have to think too much about both. >> Okay, I got to ask you, so I could wrap my stack in a container and put it into AWS or Azure or Google and it'll run great. As well, I could tap some of the underlying primitives of those respective clouds, which are different and I could run them just fine. Or/and I could create an abstraction layer that could hide those underlying primitives and then take the best of each and create an automation cloud, my own cloud. Does that resonate? Is that what you're doing architecturally? Is that a roadmap, or? >> Certainly going forward, you know, in the automation cloud. The automation cloud, we announced a great partnership or a continued partnership with Microsoft. And just Azure and our platform. We obviously take advantage of anything we can to make that great and native capabilities. And I think you're going to see in the Automation Suite us doing more and more to be in a deployment model on Azure, be more and more optimized to using those infrastructure services. So if you deploy automation suite on-prem we'll use our embedded distro then when we deploy it say on Azure, we'll use some of their higher level managed services instead of our embedded distro. And that will just give customers a better optimized experience. >> Interesting to see how that'll develop. Last question is, you know what should we expect going forward? Can you show us a little leg on on the future? >> Well, we've talked about a number of directions. This idea of semantic automation is a place where you know, you're going to, I think, continue to see things, shoots, green shoots, come up in our platform. And you know, it's somewhat of an abstract idea but the idea that the platform is just going to become semantically smarter. You know, I had to serve Re:infer as a way, we're semantically smarter now about communications data and forms of communications data. We're getting semantically smarter about documents, screens you know, so developers aren't dealing with, like, this low level stuff. They can focus on business problem and get out of having to deal with all this lower level mechanism. That is one of many areas I'm excited about, but I think that's an area you're going to see a lot from us in the next coming years. >> All right guys, hey, thanks so much for coming to theCUBE. Really appreciate you taking us through this. Awesome >> Yeah Always a pleasure. >> Platform extension. Ed. All right, keep it right there, everybody. Dave Nicholson, I will be back right after this short break from UiPath Forward5, Las Vegas. (upbeat music)

>> Announcer: Live from San Francisco, it's theCUBE covering Spark Summit 2017. Brought to you by Databricks. >> Getting close to the end of the day here at Spark Summit, but we saved the best for last I think. I'm pretty sure about that. I'm David Goad, your host here on theCUBE and we now have data scientists from Riot Games, yes, Riot Games. His name is Wesley Kerr. Wesley, thanks for joining us. >> Thanks for having me. >> What's the best money-making game at Riot Games? >> Well we only have one game. We're known for League of Legends. It came out in 2009, it has been growing and well-received by our fans since then. >> And what's your role there? It says data scientist, but what do you really do? >> So we build models to look at things like in game behavior. We build models to actually help players engage with our store and buy our content. We look at different ways we can, just, improve our player experience. >> Alright well let's talk about a little more under the hood, here. How are you deploying Spark in the game? >> So we relied on Databricks for all of our deployment. We do many different clusters. We have about 14 data scientists that work with us, each one is sort of able to manage their own clusters: spin 'em up, tear 'em down, find their data that way and work with it through Databricks. >> So what else will you cover? You had a keynote session this morning, right? >> Yep. >> Give a recap for theCUBE audience of what you talked about. >> So we talked about our efforts in player behavior where we build models and deploy models that are watching chat between players so we evaluate whether or not players are being unsportsmanlike and come up with ways to, sort of, help them curb that behavior and be more sportsmanlike in our game. >> Oh wow, unsportsmanlike. How do you define that? It's if people are being abusive? >> So what we saw was there are about one or two percent of our games that is some form of serious abuse and that comes in term of hate speech, racism, sexism, things that have no place in the game and so we want them to realize that that language is bad and they shouldn't be using it. >> It's all key word driven or are there other behaviors or things that can indicate? >> So right now it's purely based on things said in chat, but we're currently investigating other, sort of, other ways of measuring that behavior and how it occurs in game and how it could influence what people are saying. >> Maybe like tweets coming from The White House? (laughing) >> Okay, so George. >> We should be able to measure that as well. >> So how about those warriors? (laughing) >> No, George did you want to talk a little bit more >> Sure. >> David: about the technical achievements here? When you look at like trying to measure engagement and sort of maybe it sounds like converting high engagement to store purchases, tell us a little more maybe how that works. >> So we look at, we want. Our game is completely free to play. Players can download, play it all the way through and we really try to create a very engaging game that they want to come back and they want to play and then everything they can buy in the store is actually just cosmetics. So we really hope to build content that our players love and are happy to spend money on. As far as... We just really want engagement to be from around players coming back and playing and having a good time and it's less about how to get that high engagement conversion into monetization as we've seen that players who are happy and loving the game are happy to spend their money. >> So tell us more about how you build some of these models like, you know, turning it into not turning it into Spark code, but how do you analyze it and, sort of, what's the database mechanism for, you know, 'cause the storage layer in Spark, you know, is just like the file system? >> Sure, yeah absolutely. So we are a world-wide game. We're played by over 100 million players around the world >> David: Wow. >> And so that data comes flowing in from all around the world into our centralized data warehouse. That data warehouse has gameplay data so we know how you did in game. It also has time series events, so things that occurred in each game. And our game is really session based so players can come play for an hour, that's one game, and then they leave and come back and play again. And so what we're able to do is then, sort of, look at those models and how they did. And I'll give you an example around our content recommendations. So we look at the champions that you've been playing recently to predict which champions you are likely to play next. And that we can actually just query the database, start building our collaborative filtering models on top of it, and then recommend champions that you may not play now, you may be interested in playing, or we may decide to give you a special discount on a champion if we think it'll resonate well with you. >> And in this case, just to be clear, the champions you're talking about are other players, not models? >> It's actually the in-game avatar. So it's the champion that they play. So we have 130 unique champions and each game you choose which champion you want to play and so then that plays out for like. It's much more like a sport than it is like a game. So it's five v five, online competitive. So there are different objectives on the map. You work with your team to complete those objectives and beat the other team. So we like to think of it like basketball, but with magic and in a virtual world. >> And the teams stay together? Or are they constantly recombining? >> They can disband, yeah. Your next game may find nine other people. If you're playing with your friends then you can just keep queuing up with them as well. So the champions that they control there happen to be who you're playing in that game. >> And when you are trying to anticipate champions that someone might play in the future, what are the variables that you're trying to guess and how long did it take you to build those models? >> Yeah, it's a good question. Right now we are able to sort of leverage the power of our user, our players, so we have 100 million. And so what we do and we have in our game there are roles so, for instance, like there's a center in basketball, we have a bot lane. So we have bottom lane support and bottom lane ADC. So a support character is there to make sure that your ADC is able to defeat the other team. And if you play a lot of support, odds are there are other players in the world who play a lot of support too so we find similar players. We find that if they engaged on the same sorts of champions that you play. For instance, I'm a Leona main and so I play her a lot. And if I were to look at what other people played in addition to Leona it could be things like Braum and so then we would recommend Braum as a champion that you should try out that you've maybe not played yet. >> David: Okay. >> So and then what's the data warehouse that you guys use for the ultimate repository of all this? >> All the data flows into a Hive data warehouse, stored in S3. We have two different ways of interacting with it. One, we can run queries against Hive. It tends to be a bit slower for our use cases. And then our data scientists tend to access that all that data through Databricks and Spark. And it runs much quicker for our use cases. >> Do you take what's in S3 and put it into a parquet format to accelerate? >> Sometimes, so we do some of those rewrites. We do a lot of our secondary ETLs where we're just joining across multiple tables and writing back out. We'll optimize those for our Spark use cases and there's writing back, sort of, read from S3, do some transformations, write back to S3. >> And how latency-sensitive is this? Are you guys trying to make decisions as the player moves along in his level or? >> So historically we've been batch. We do- our recommendations are updated weekly so we haven't needed a much higher cadence. But we're moving to a point where I want to see us be able to actually make recommendations on the client and do it immediately after you've finished a game with, say, Leona, here's an offer for Braum. Go check it out, give it a try in your next game. >> So Wesley what would you like to see developed that hasn't been developed yet that would really help in your business specifically? >> So one thing that's really exciting for gaming right now is procedural generation and artificial intelligence. So here there are a lot of opportunities, you've seen some collaborations between Deep Mind and Blizzard where they're learning to play Starcraft. For me, I think there's a similar world where we have a game that has different sorts of mechanics. So we have a large social piece to our game and teamwork is required. And so understanding how we can leverage that and help influence the future of artificial intelligence is something that I want to see us be able to do. >> Did you talk with anybody here at the Spark Summit about that? >> Anyone who would listen. (laughing) So we chatted some with the teams up at Blizzard and Twitch about some of the things they're doing for natural language as well. >> Alright so what was the most useful conversation you had here at the summit? >> The most useful one that I had, I think, was with the Databricks team. So at the end of my keynote, It was kind of serendipitous, I was talking about some work we had done with deep learning and sort of doing hyper parameter searches over our worker nodes, so actually being able to quickly try out many different models. And in the announcement that morning before my keynote, Tim talked about how they actually have deep learning pipelines now and it was based on a conversation we had had so I was very excited to see it come to fruition and now is open source and we can leverage it. >> Awesome, well, we're up against a hard break here. >> Wesley: Okay. >> We're almost at the end of the day. Wesley, it's been a riot talking to you. We really appreciate it and thank you for coming on the show and sharing your knowledge. >> Wesley: You bet, thanks for having me. >> Alright and that's it, we're going to wrap it up today. We have a wrap-up coming up, as a matter of fact, in just a few minutes. My name is David Goad. You're watching theCUBE at Spark Summit. (upbeat music)