 Live from Washington DC, it's theCUBE, covering Boomi World 19, brought to you by Boomi. Welcome back to theCUBE's coverage of Boomi World 2019 from DC. I'm Lisa Martin, John Furrier is my co-host, and we're pleased to welcome a couple of guests from Accenture, Boomi Partner. To my right, we've got Pete Yao, Global Managing Director of Integration, and Leticia Kayatu, Accenture's Global Lead for Conversational AI. Welcome both of you. Thank you, it's great to be here. So big news, you can't go anywhere these days without talking about AI. I mean, there's even commercials on TV that any generation knows something about AI. But Leticia, let's start with you, some big news coming out this morning with what Boomi and Accenture are doing for conversational AI. Give our audience kind of an overview of what you guys announced this morning. So thank you very much. So conversational AI is booming in the market. It's at the top of the agenda for a number of our C-suite. It's a new way to make system more human. So instead of having to learn the system, you can actually speak, ask them direct question, and have a conversation. And actually what we're doing, what we've announced this morning is Accenture and Boomi are going to partner together to deliver that kind of services for our clients much faster. Because we have the expertise on the know-how of designing those conversational experience. And Boomi obviously integrates really fast with backend system. And the two together can really be accelerating the value delivered to our client. And the technology piece, just want to make sure it's in, because you guys are providing a front end, so real technology with Boomi. So it's a together story? Yeah, it's definitely a together story. And as you say, we are quite experts in designing those experience on the front end. And Boomi obviously kind of powers up the integration in the background. So this is going to be enabler of, something you said a minute ago, is instead of us humans having to learn the tech, the tech's going to learn us, is that fair to say? Very fair to say, that's exactly how we want to see it. And I think we call that trend, radically human systems. So systems are going to become more radically human as we go on. And the conversational layer is one enabler of that. Is it going to be empathetic? Like when you're saying this morning, something I love on stage, that we've all had these interactions with AI, with bots, whether we're on a dot com site, trying to fix something for our cable provider, or we're calling into a call center, and you're starting to get your voice changes, your agent, and you want that. Is it going to be able to understand, oh, all right, this person, maybe we need to escalate this, there's anger coming through the voice. Is it going to get to be able to detect that? On voice you can definitely start detecting tone much better than on text, because on text is very small snippets and it's quite difficult to define somebody's mood by one small interaction. Typically you need a number of interaction to kind of see the build up of the person's emotion, but on voice definitely, you know, your intonation definitely defines your state of communication, yeah? You can tell if someone's happy, sad, and then use the text metadata to add to it. This is fascinating, because we all see Apple with Siri front end. That's a different system, they have a backend to Apple, this is a similar thing. You guys have a solution at Ascension, can you explain how people engage with Ascension, because the Boomi story is a great announcement, congratulations on that, but still, you can deploy this technology to any backend, is that right? To any backend, we have a number of live deployment running at the moment. I think the key thing is, especially in the call center, call center is an area that has not been invested in for decades, yeah? And very often, the scripts are very inward driven, so they would describe the company's processes, rather than think about the end user. So what we do in Accenture is, we try to reinvent the experience, be much more user driven, and then we have a low code, no code kind of interface to be able to craft some of those conversations on all their variations, but more importantly, we actually store all those conversations and can learn, so we have assisted learning module to make your natural language processor cleverer and cleverer. And as you were saying, before we started to be on here, the user is contributing training data. Yeah, I was just sharing that, one of my recent stories of an ISP that I was trying to interact with, and frustrated that I couldn't just solve this problem on my own. And then after I was doing some work for theCUBE a few months ago, I realized, oh, actually, I have to be calm here. I have an opportunity, as is everybody, to help train the models, because that's what they need, right? It takes a tremendous amount of training data before our voices can become like fingerprints. So I think if more of us just kind of maybe flip that, maybe our tone will get better and obviously the machines will detect that, right? Yeah, I know, definitely. I think the key with conversational layer is not to see it as just pure tech, but really an opportunity to be more human centered, and obviously knowing who people's on, how they interact in different kind of problem and scenario is absolutely critical. Pete, I want to get your thoughts on digital transformation because, you know, we've been, I mean, we've done, I've done thousands of interviews on theCUBE and many, many shows. Digital transformation's been around for a while. It all stops in one area, okay? Process technology, great areas, we've got visibility in that. Automation's excellent for processes, technology, a plethora of activity. The people equation's always broken down. Culture has stopped DevOps, maybe not enough data scientists or linguistic engineers to do conversational AI. You guys fill that void, great technology. The people equation changes when there's successes. It all comes down to integration because that's where, you know, either I don't believe in it, I don't want to do it, the culture doesn't want it, time to value. The integration piece is critical. Can you guys explain how this boomy center integration works and what should enterprises take away from this? Well, yeah, one of the key things when we started our relationship with boomy more than five years ago now, really, boomy was the leader, kind of the ones who invented the iPass, right? The integration platform as a service. So in the small and medium business, a lot of those companies had already moved a lot of the critical apps to the cloud. But in the enterprise, we see that it's taken a lot longer, right? So certain departments may move certain pieces, but it's still very much a hybrid, right? Between cloud and on-prem based. So taking a platform like boomy and being able to use that with the AtomSphere platform has really allowed us to move forward. We've done quite a bit of work in Europe and now in the last year we've been focusing in North America along with Europe. So really the platform has allowed us to focus on the integration. It's interesting, you guys both have been in center for a long time, you've seen the waves, you know. Oh, big 18 month deployment, eight years. I mean, sometimes years go back in the 80s, 90s. But now the large enterprise kind of looks like a lot of SMBs. Because the projects all look, they're different now. You can have a lot of projects out there, hundreds of projects, not one monolith. So this seems to be a trend. You guys see it that way, do you agree? Could you like share some insight into what's going on in these large companies? Is it still the same game of a lot of big projects or are things being broken down into smaller chunks with cloud platform? You guys just share your insights on this. You can do first, yeah? Okay, so the days of the big bang, big transformation, multi-year programs, we don't see very many of those, right? A lot of our clients have moved away towards lean, agile delivery, right? So it's really being able to deliver value in shorter periods of time. And in that sense you do see these big companies acting more like SMBs, right? Because you really have to deliver that value. And with Bumi's platform, it's not just the integration aspect and though our relationship started there, it's with some of the other pieces to the technology, like Flow and Low-Code or No-Code as well, which has allowed Bumi customers and our clients and our teams to be able to get those applications out to production much quicker. Yeah, big name blur of, sorry, of the citizen developer. Yeah, absolutely. Thoughts on this trend? Yeah, so I guess my thought, I will come with the innovation angle. So obviously we're in a very turbulent time where a company, like a number of the FTSE 500 of 20 years ago, they're not there any longer and there's quite a heavy rotation on some of the big cooperation. And what's really important is to size the market and innovate all the time. And I think that's one of the reasons why we have a much smaller project. Because if you want to innovate, you need to go to market really fast, try things up and pivot the new ideas really fast to try to see if people like it and want it. And I think that's also one of the key driver of smaller kind of projects that would just go much faster to life. We had a guy on theCUBE say, data is the new software. Kind of, you know, bringing a provocative statement around, data is now part of the programmatic element and integration speaks volume here. So I want to get your reaction to the idea of glue layers. I mean, people kick that turn around. It's a glue layer, basically integration layer with data, control plane. This is a really big part of the integration story for Bumi, but for other customers. What's your guys thoughts on, you know, this data layer, glue layers as software and data come together? You're showing it with the conversational AI, it's voice, turns to software, connects to another system, there's glue. Yeah, so that's a very interesting angle. Because I think, you know, in the old integration world, people would just build an interface and then it would go live and they wouldn't necessarily know exactly what's going on the bonnet. And I think having that insight of what's your flow or often they use when they're kicked off is something that becomes quite important when you have a lot of integration to manage. I would remember I was working for a bank, a major bank in the UK, where we were trying to, you know, make a mainframe system go real time. But we had all those batch schedule kind of running and nobody really knew when what and the dependency in between each other. So I think it definitely helps a lot, you know, bubbling up that level of visibility you need to transform truly. Yeah, and you're seeing a lot of companies now have chief data officers, right? But data really is important. And with big data, data lakes, unstructured data, structured data, traditional RDMS databases, being able to access that information, is it just read only, is it read and write? You're really seeing kind of how all of that has to come together. So if we look at the go to market for booming Accenture, Pete, talk to us about how that go to market strategy has evolved during the partnership and where you see it going with respect to, you know, emerging technologies like conversational AI. Oh yeah, we've got great opportunities, right? So we started off really just, hey, there was an integration opportunity and you know, are we doing much work with Boomi and the Enterprise? Five years ago, we hadn't, right? And we started doing more work kind of in Asia pack and then in Europe, three years ago, we entered a formal relationship to accelerate the growth. It was accelerated growth platform, which started in the MIA. And this last year, we formally signed one in North America as well. And in the last three years, we've done four times the amount of work. The number of customers, we've got more than 40 joint customers together. The number of trained professionals within Accenture, we have more than 400 people certified with more than 600 certifications. Some of them may be a developer as well as an architect. And so a lot of that is really that awareness and education, training and enablement, as well as some joint go-to-market activities. Any of those in a specific, I was reading some use cases in healthcare and utilities. Yeah, we're definitely, we've seen quite a bit in utilities in our energy practice. We've seen it in transportation. We've really, because Accenture covers all of the different industry groups, I mean, we're really seeing it in all of them. You know, I'm fascinated by this announcement you guys had with Bumi, the big news, conversational AI, because it just makes so much sense. But I worry people will pigeonhole this into, you know, voice, like telephone call centers only. Because the use cases that you guys were showing on stage was essentially like almost like a query engine and using voices versus like agent, you know, call center workflow, which is a natural workflow, big market there. No doubt about it. But there's other use cases. This is a big, wide topic. Can you just share the vision of conversational AI a little further? So, meaning, I think the capability we have is to kind of be able to go on any channel. Voice is an interesting one, because it's, I think it's very common still, you know, to have a call center when you're deep into challenges. And this is kind of the most emerging and challenging from a technology perspective. So that's the one that I showcased. But there's a number of chat channels that are also very important on the web or a synchronous channel like WhatsApp and Facebook and all of that kind of thing. So it's really kind of really offering a broad choice to the end consumer, so they can pick and choose what they want at the moment they want. I think what we see in the market is a big shift from synchronous kind of interaction, like on the web. You go on the web, you chat with something and you just need to be there to finish it. To actually text, because you can just send a text, get a response, go to a meeting and on the back of the meeting when you have five minutes you just kind of do the reply. And you actually solve your problem on your own terms, but really when you have the time. So there is a lot coming there. And you know, with Apple business chat, you know, there's a number of mechanisms that are coming up on new channels. Before company tended to be, you know, we do digital, we do call center and maybe we have, you know, chat, but actually all of that is broadening up. You know, people want multi-channel experience, yeah. Yeah. So a synchronous is key, synchronous and asynchronous communication. Is there a tell sign for a client that says I'm ready for conversational AI? Is there a way I have to have a certain data set? I mean, is it an interface? What are some of the requirements? Someone says, hey, I really want this. I want to do this. Yeah, so the way we deal with all of that very often is if you have call center recording or chat recording, we have a set of routines that we pass through. So we transcribe everything and we do what we call intent discovery. And from that, we can know, you know, what are the most kind of critical kind of processes kicked off. And from that, we know if it's transactional or if it's an interaction or an intent that's emotionally loaded, like people not happy with the bill. And then we have different techniques to address all of those different kind of processes and if you want to transform them into new experiences. And we can very easily kind of look at the potential value we can get out of it. So for instance, with one of our clients, we identify, you know, if you do that kind of transformation, you can get 25 million of your call center. You know, like, which is very sizable and it's very precise because it's data-driven. So it's based on real kind of calls, recordings and data. Can't hide from data. I mean, is either successful or not. You can't hide anymore. Yeah, and I think one of the extra value add is very often calls center agent or chat agent, they're not really paid to classify properly. So they would just, you know, pick up the most easy one all the time. So they wouldn't misclassify some of those recordings to, you know, what's easiest for them. But when you actually go into what was said, it's a very different story. Great insight. AI becoming not just IQ, but EQ in the future? Yes, definitely. That's the old idea. That's why we need our users to embrace it. Exactly, and how, and turn those frustrating experiences into I have the opportunity to influence the model. Last question for you in terms of conversational AI and the business opportunities that this partnership with Boomi is going to give to you guys at Accenture. Oh, definitely looking forward to joint credit market taking this globally. We were named earlier this week, yesterday, the Worldwide Partner of the Year. Second time that Accenture has been awarded that, which we appreciate, and that we look forward to working with Boomi and taking conversational AI to our joint clients. Awesome, Laticia, Pete, thank you so much for joining John and me, really interesting conversation. Can't wait to see where it goes. Great, thank you very much. Our pleasure. Very conversational. Very conversational. Guess the AI here, come on. This time we give you a bot to sit in on. Yes! Huge conversations. Exactly, for our guest and for John Furrier. I'm Lisa Martin, you're watching theCUBE from Boomi World 19. Thanks for watching.