The Rise of Enterprise Chatbots: One Bot for all your Apps


Transcript: So, this will be a little different I think because Oracle as you probably know the Enterprise company, my name's Carlos came with I'm in marketing side. I'm al so, joined by Giovanni who's in the project management side the outbound product management side. And it's a corporate slide basically says I have to present this every time but basically says don't take any of this as a promise and don't make any rash business decisions based on what I say. Oh, you're doing it. All right. All right. so, guess I wanted the first start with saying that we definitely live in interesting times and by interesting times. I mean that I don't mean that as in the you've probably heard the Chinese curse supposed Chinese curse of may you live in interesting times, right which by the way doesn't actually it's not a Chinese curse. Someone made that up look it up Wikipedia before this before I said this just because I Be thorough in this what I meant was, you know, the past 20 years you've least I have many of you probably have as well have seen this huge transcendental Trends huge changes for the world for the for our world. Right? We had the World Wide Web which connected millions of computers and then we had the iPhone moment what I like to call the iPhone moment is when you're effectively taking all those computers really what's in your pocket the phones that you're looking at right now. Are really just computers disguise. This phone’s so, those computers that were connected now are on your iPhone or your Android device Etc. Right? And then later on another as time went by those computers. A lot of those computers shifted up to the cloud and the shifting up to the cloud gave massive amount of amounts of power massive Amin basically infinite amount of storage which allowed AI to prosper which allowed ai scientists to take advantage on cheap storage. Cheap power to have these huge benefits from AI so, the window to AI lot the popular window to AI are going to be chatbots. So, that's why from this timeline. We're seeing everybody's got to have a website that was 20 years ago. Everybody needs a mobile app that was about ten years ago and the launch of 2007. I think it was Steve Jobs it done and then now it's more and more. You know, everybody's talking about Alexa Google home Etc. And now those chatbots. Because we are all consumers those chatbots are moving towards the Enterprise, which is our main topic here. Oops. Oh so, this is the gardener quote that's by 70 by 20 20 20 to 70 percent of white collar workers. Probably most of us here will be interacting with some sort of conversational interface. Click and that obviously becomes very natural for us. Right because probably most of this audience kind of knows this already but in case you don't know or you need to be briefed a little bit more on this natural because conversational interfaces, there is no interface right what you need to interface with a conversational interface. You learn in grade school, right? You learn how to talk. You don't ask questions. You learn to ask for cookies or even type or even at least little baby little older learned how to type and then The barriers to entry for a chatbot are so, much lower right there. You don't have to go to a website. You don't have to download a mobile app. You're just simply using text messaging or the slack or the Microsoft teams, whatever text messaging interface that you're using to get to it. so, there's nothing to download. And of course, chat Bots are available 24/7 all the time. It's a new channel. Everybody wants to be on it and ideally in a perfect world. You have a chat bot that interfaces all chat Bots, which is what we're hoping to present. I think it was a Doe B that was talking earlier about call centers. I thought that was really good. It's a Strong use case for us a lot of our customers use that naturally high call vole of chat bots are perfect for that. Right? When you have a high call vole you were able to address wait times you are able to address customers immediate needs and a lot people nowadays, especially the younger generation don't even want to talk to a person right? They just want to go take care of business on a text and you're done right and Staffing Resources, you won't have to staff as many people or the Staffing research that you do have can actually work on more complex problems as opposed to Answering what's the status of my delivery things like that? Right bring it on more to the Enterprise applications whether you're using a CM which stands for human Capital Management basically for HR your enterprise resource planning that can systems again as well as CRM, you know chatbots being the next big thing. mean who wouldn't in HCM and HCM land want a chatbot that makes employees more productive being able to automate recruitment HR would love to on as much of the recruitment as well as much as much of the onboarding as possible. Lots common use cases their improved access for Erp systems, you know back in systems like that. You really need to jump on the web jump onto the web application find out what's going on with your supply chain when we better if the chat bot was proactive and be able to text you message saying hey somebody just raise the tariffs across the world you're going to have problem to get information that way sent to You that's wrong for you as opposed to you know, navigating to a website or using a mobile app. On the CRM side. I mean sales folks who doesn't want those folks always want as much information about their customers as possible for making a call who doesn't want a more enabled sales force to help shorten the sales cycles and accelerate those sales. so, you can mobile apps are still fantastic. Don't get me wrong Chiapas. You're not going to replace anything every one of us use the mobile app today probably to figure out where we're going to eat, you know got an Uber over here or left over here or you know, trying to figure out what to do in San Francisco after the day, right? But the shortage with the shortcomings of mobile is that there's a limited click. There's a limited real estate that you're going put up. Right. There's only a couple dozen apps that you were I would argue that you probably use regularly right in that same vein chaplets have that same challenge, right? Do you read click? sorry. Do you really want to have to interact with one pot here one about their one about everywhere and repeat those same mistakes. You're going to have challenge that you've had with mobile apps click and I said it was you and so, to that end. What we wanted to do in the this is a little this presentation little bit of a challenge because we want to do live demos for Live demos with back-end systems on our servers or computers over there. so, I'm going to have to run over there and do it really quick. But knock on wood, it'll all work. I think the demos comes up next right? Right. so, a true story. We just had a conference right and work. Oh, and Carlos here when I was talking, we had a situation where he needs to hold a mic for me for extended period time. so, now I'm going to return the favor and hold the mic for Carlos. Thanks. Thank you. Yeah, that is true story. had to hold the mic is guess we got a selfie of that too. so, let me start the story with our sales person who's using our chat pod and Enterprise chat pod. Just this one the one that's going to close this really quick. Well this down. All right. so, an Enterprise chat pod and this is again live hitting our servers and she's a salesperson and wants to go okay with salesperson was my quota so, you can see there's a quota involved. so, now the salesperson is effectively hitting CRM system real back in real CRM system to Practical political order they said okay. What are my opportunities? Okay A little delay, so, I'm not going to put a shorten this a little bit. so, there's a lot of opportunities going on. And now the salesperson says okay, you know what we've gone through a reorg. I don't even know who my manager is anymore. Just gave my manager. so, the Nuance here is that we're Switching gears between a CRM system and ACM system. human Capital management system, right? so, I'm finding out who my manager is and I'm very confident now. Okay. I've got this big deal coming down and know what's coming down the pike. I'm going to make a lot of money. First person. I'm going to plan my vacation too in this case Africa, but go to Africa and I know I got to go to go to Africa. I’ve got to go get my shots done and all those kind of things as you go into 2/3 to Africa. so, you have to get that done. so, in this case from an HTM application, I need to go to the hospital get the shots done and I don't know what my medical group ideas. So, I'm just going to say my medical Heidi and sure enough it'll come back with medical buddy. I can go get my shots. Great. Thanks bot. Alright, so, I come back out you my business. so, we've touched already on a CRM application and HTM application now I go back out to the car. someone broke into San Francisco little tough rough neighborhood. so, in this case, I'm going to say oh no big deal five minutes already, but no big deal. I'm going to order new laptop order. New laptop you can type, sorry. Yeah, this is the old one and say yes. This is asking me about my role. What kind of laptop second order so, it's going to give you bunch of out that I'm going to cut to the chase here real quick and pack go past that what I what I do want to say here in just finish off here is that we're hitting a CRM system and HTM system one Erp system back at separate back ends from one pot. so, you don't have to switch gears all the time. And that's what we refer to as our Enterprise digital assistant. so, here I'll switch probably hold the mic for a Joe here, okay. so, thank you. I was told don't have lot of time. so, I'm going to get to it and during open world which happened last week. We also, announced the Oracle voice SDK Oracle voice. Now a lot of you may be wondering why are we doing voice? Right and mean there's Google and Amazon do a great job in voice recognition. Why is Oracle coming out with a voice recognition engine to do that? Well, first of all, our customers are telling us well, we're concerned.


With the Privacy especially with Enterprise data going through Alexa. We're going through a Google I forgot, you know, my apologies at whether it's Amazon Google they had the team of people behind the scenes listening to we don't have that. Right. so, privacy is important to us. That's one of the key tenant that why we're doing that GDP are on Minos gdpr is more of a European kind of thing data privacy, right? We're complying with that. We want to make sure that the that's complying with that.


Another reason and this is actually a very important reason is with Enterprise data and usage of the chappa'ai. A lot of terminologies are specific to Enterprise. so, for example name and account or even a terminology whose honeywell's CAD, if you know somebody worked for Oracle or a large account where large companies they know that means key account director, but if you ask a luck so, Google that its chances are going to be translated into Honey's CAD. Well and name busy different and the reason so, the reason we have our voices. K is so, that we can train these Enterprise specific data and terminology. But al so, we allow our customers to do a better recognition of the names. so, let me switch over to do a quick demo and this is going to be a really interesting demo because oh yeah, where is that?


Okay because as you know, we're I'm doing a demo through a laptop microphone, but I'm talking through microphone. so, you have multiple level of interference Sure, everyone can see it. Okay going to give a shot. Alright, so, the scenario is this I'll do a very quick because it's a Voice demo and go very quickly the scenario is basically a salesperson, they before trying to prepare to visit a site. so, let's say I'm salesperson 11 a.m. Okay, let me start over. Send me the details of the last two meetings. Alright, so, we did that. Do we have any reference will customers using autonomous database in retail? Yes database. Yeah, so, as you could see it. We already trained it to understand work will products. And of course, we

can allow customers to try and understand Oracle product.


Now, what about specific names? Who did we compete with four Elemental design? And what was the deal size database who are reference a rule? What account does Bob have except Elemento design that use autonomous database? Right. so, this is one of the things that we're also, working on is you know, how many you know that we can actually process not only positive but al so, it's very difficult for typical chatbot to understand negative terms except and this is something that we're able to do accomplish. And me see. Let me do two more quick thing. Send a wind report for the second one to Chris.


Then didn't quite work correctly. As I said, this is a very interesting with laptops going through a microphone. All right. so, the next one is really going to be interesting the morning. I need to get more information. so, I need to go to a Starbucks. so, let's say find a Starbucks closest to fancy cosmetics and send the address to Sangha Viswanathan in Michael Kennewick. Here's the information only Starbucks. so, it's able to identify the Starbucks. Anyway, I'm going to cut the demo short because when running out of time and al so,, this is a very cool demo. I don't know. so, this is the concert just took place last week at Oracle OpenWorld.


You can't hear whatthe guy is saying, right? The guy is trying to do some work in the middle the concert. And this is a little test that we're doing and this is another reason why we're developing our own voice SDK. so, let me show you what person I guess. so, this is the with the noise. He's kind of filter out little bit we can actually hear what's going on.


So, the other thing again, what we're doing is the reason we developed this technology is to help filter a lot of these noises and make the voice recognition lot more accurate. so, that's what we're doing. We're doing voice is running security privacy. We al so, have a large ecosystem and is open. Okay, we'll just do the one last one. so, just quick thing three summary one is chatbot.


They're coming to Enterprise to is Enterprise is different from the consumer space and for us trust and security is a key requirement and last thing we recently

announced partnership with Microsoft. so, Oracle and Microsoft actually networking together with Microsoft teams. so, we have now tight integration with Microsoft teams. How many of you use Is Microsoft team and Enterprise settings? Okay, a few we have a lot of customers that actually does that. so, this allows us to do deep integration with Microsoft team. Al so, allow Microsoft team to have a conversation use as a conversational client into all of our SAS applications are applications.


Okay, Round of Applause, please. We got time for one quick question. Okay. The question I have is did you actually give two commands to your Oracle bot that's amazing to command for the voice part or yeah, the boys plug. Yes. Yeah, we are able to do that because the other thing that we found out is when you use voice interact with a chatbot, you know is a human nature to have much more complex utterance you type, you know, okay type a single sentence, but for voice you tend to have more complex structure. so, that's one of the things that we're working on is an advanced nlu or NLP understanding complex. Utterances and multi intent resolution and this term semantic parsing and pure you hear a lot. But that's part sound that data science test. We have that are going about to solve that. Thank you very much. Thank you. Thank you, Carlos.

© 2017 by Chatbots Conferene TM.