I'm very pleased to be able to introduce our lead sponsors, Kovio, who are going to be joining us now, both to talk about the future of generative AI in CX and what they are seeing from their perspective, but also to give us a bit of a demo as well in this half an hour. So it's gonna be really quite hands on. I'm very pleased to welcome both Bonnie Chase, who is the senior director of Service Product Marketing at Cobio, and also Brian Chancellor, the sales engineer of Cobio as well. Bonnie, I can see that you're already joining us. A great to see you. Thank you very much indeed. Hello. Thanks so much for having us. More than welcome. Okay. So, Bonnie, I can see you've got some slides. So, what we'll do is I'll hand over to you now so that you can introduce yourself to the audience and talk us through a little bit of an introduction from your side. And then, yeah, we can catch up with Brian and Little afterwards. Awesome. Sounds good. Thank you so much. Hello, everyone. Thanks so much for joining our session. As Georgina said, I'm Bonnie Chase. I lead the product marketing for, our service solutions at Koveo. And joined today by Brian Chancellor who'll be doing a demo at the end. Hey, Brian. Good morning. Good morning. Alright. Now before we get started, I do want to, to share a forward looking statement because we do wanna share some exciting product developments with you today while we have you. Don't worry. I won't be reading this to you on the out loud on this session. But it will be available, when we shared shared the re recording out. Now for those of you who are not familiar with us, COveo has been building enterprise ready AI for over a decade. And our goal really is to democratize AI so that any prize can take advantage of AI capabilities for search recommendations and one to one personalization. And we do this over three lines of business commerce, service, and platform, which includes both website and workplace. And ultimately, the goal is to deliver those personalized, relevant, and profitable experiences across the entire customer journey, which is why we're here today. You know, today, we're talking about generative AI and and to us, best of breedings were constantly pushing the envelope and and staying ahead of the technology curve. So we've been doing AI for over a decade, generative AI is just the latest of development in our our journey. So I wanna jump in by starting with a fun, story. So if if you don't recognize this picture, this is this is in banff. And I got to go on a little big in a couple of weeks ago to, you know, go on a bamp, do some hiking, have some fun. And before doing so, I wanted to purchase a gimbal. And this is really, you know, a tool to help you stabilize your camera, you know, for recording purposes. And as I was shopping, I had specific technical requirements of what I needed. I wanted the stabilizer I wanted it to be foldable and compact, have facial recognition, all of these different, you know, technical capabilities. Right? And at the end of the day, my outcome was to create a video to share with family on social so they could get that experience. Now when I got the Gimbal, and I was ready to test it out. It had all the technical requirements that I asked for, but the image stabilization only worked in landscape mode. Not portrait, which means if you want something on social, it's not really gonna work that well. So at the end of the day, I I thought I had all the requirements. I made the purchase, and it did not create the experience that I wanted it to create. Now why am I sharing this with you today? Well, ultimately, when we look at b to b purchases, b to b buying, processes, it's not that different than what a consumer goes through. And there are really three key considerations when you wanna make a successful purchase. One is the outcomes. You know, what outcomes are you trying to achieve? What is the experience that you wanna create? Then what is the technology. And oftentimes when we're making these purchasing decisions, we we start with the technology without thinking through the experience we wanna create. That drives the outcomes we're trying to achieve. And when it comes to B2B software, and when we think about this, ultimately, what needs to happen is the delivery needs to be in a connected total experience. So remember that chatbot revolution that we were promised I really liked this quote here. You know, I'm not even sure if we can say chatbots are dead because I don't even know if they were ever alive. So this is, you know, we we went through this before. We we had new technology come out. We focused on the technology without thinking through the experience we wanted to provide, the outcomes we wanted to achieve with it, and then ultimately it was not delivered in a connected way across the total experience. And so to no surprise, the chatbot revolution did not cannot the way that we expected it to. And this is something for us to remember because we wanna make sure that we're not making the same mistake with generative AI that we made with chatbots, putting all your eggs in one basket, not really thinking through that connected total experience. Now if we think about what defines a total experience, this is a definition by Gartner, and it's really a strategy That creates that superior shared experience by interlinking all of these different components. You have the multi experience, which is really about providing that simpler interface across multiple touch points, the user experience, making sure that you're providing that effortless more intuitive you know, navigation, UI experience across multiple devices. The customer experience, which you know, it's around increasing that engagement. And then, of course, the employee experience, which is around increasing empowerment. And what we found is that, you know, total experience, if you're focused on that and you're providing that total experience, It will help you outperform competitors by twenty five percent. And this is a significant number because, you know, total experience is important because it really allows companies to think holistically and hopefully achieve economies of scale by being able to cater to various audiences with fewer tools and processes. So total experience is about breaking down silos and empowering teams to get more not necessarily do more, but get more with less. Now when we think about generative AI and for us, you know, at Cobail. When we're talking about generative AI, we're really focusing in on the generative answers part. So you do a search you get an answer provided by generative AI. But if you think about the entire user journey, and how important it is to have, you know, unified answers to have the same source of information providing those answers across the experience it's important that generative answering does the same. So, you know, you don't wanna put one search box in, you know, the presales experience and then generative answering and the post sales experience and then the per and having to be confused about which search they need to use, you know, which one has more accurate information. When you look at the full journey. What you can see is that it really is, knowledge that pushes people through every stage. So whether you're looking for something to buy your browsing all the different tech technological requirements, whether you are a new customer, you're onboarding, you're trying to learn the product, or maybe you're an existing customer who wants to use learn how to a product even better. At the end of the day, it's unified knowledge and content that helps people get the information that they need to move on to the next step in the journey. And the idea is that it shouldn't matter where your users are along their journey or which sites they are visiting, they should have a seamless secure experience across the board, and that doesn't change with generative AI or generative answering. That's something to think about, you know, we see GenAI not as a separate siloed interaction channel. Know, search is not going away. We're still leveraging the same data, the same user, the same context. The answers are just provided in a different way. So as we think through this and, you know, as we looked at a few different generative answering solutions and and how we wanna to approach this. You know, Koveo search is built on a, using relevance on a unified index. And so this is fed with secure connectors on the entire enterprise content. So you can see, you know, in this example, this is just a basic search. Right? You get all of your enterprise content push through secure connectors into a single index, and that relevance is if that a those AI capabilities are applied to sure that you're getting the best results. Now when you think about, you know, the new technology with NAI, question answering large language models, It is possible to build a system that will process knowledge based content and then, organize it inside a vector database to then be combined by a large language model to generate the answer. But the problem is that you get two different search boxes, two different you know, content sources and different set of facts in each of these experience. And that's that's not what you wanna provide for your customers. Right? So that those different search boxes create confusion, the duplicate content, you start losing trust because you're not sure where to search, what is factual, and where the information is coming from. So as we think through you know, having an integrated search and question question answering solution with generative answering, we're really thinking about search as a new omnichannel, you know, component. It's it's something that should be the same across the board. Instead of two separate experiences through one search box, It's really about creating a unified experience. So just to walk you through how we're doing this from a high level perspective, What we're really looking at is from the back end, making sure that you have the depth and breadth of content you know, the freshness of content. So anytime you're making an update, no matter where that update to content is happening, whether it's in Salesforce knowledge or confluence, you wanna make sure that that's automatically indexed. Taking in the security and permissions, as well as having the administration capabilities and the analytics behind the performance of those generative AI capabilities. On the front end, you get a few different things. Right? You get the unified search box for all the queries. You get the generated answer on the most relevant paragraphs, personalization, and then protection against hallucination because it's based on on grounded context. Text. So again, the idea here is that rather than creating, you know, multiple search experiences, you know, maybe a search here and generative answering here, you wanna be able to layer that into every touch point that you have and a platform that can allow you to do that, is really important to make sure that it's getting all of that content and being able to be extended into each of those touch points. So as an example, and, you know, in Brian's demo, he's gonna walk through in more detail, but Here's an example of combining search and answers. You know, you have your normal search, you have your generated answer at the top, and then you have those search results that feed into that generated answer. Now what's really important when you think through, you know, what are some of these risks for generative AI, you know, hallucinations is one of them. You know, there are a lot of people are worried about making sure that we're providing accurate information that our customers can trust. And with that, you know, it's important to include sources and citations. So with Covayo, what we're doing is we're making sure that whatever content is leveraged to generate that answer, you can actually validate that because we'll include those sources and citations right there within the answer. And then in the future, you know, we're really looking at continuing to evolve. With follow ups, conversations, being able to reformulate those questions, really making it more conversational so that it's easier to drill down into the specific answer. But the key is that all of this is based on the same content with the security and permissions built in. To give you an example, you know, we have gone live in multiple areas of our own customer journey So we have generative answering in our documentation site within our community and within the in product experience. So as somebody is you know, using our product and they have a question, they can get that answer right from within the product. And all of this once again, is leveraging the same content. It's delivering the same experience, with permissions in mind. So, obviously, you're you're only seeing, content that you should be able to see. So at the end of the day, you know, what this really means and and what what the objective is is you wanna take that all of that content and data, right, that you have, that you wanna be able to generate those answers on, really focus in on what those are. Pushing that through a relevance engine like Coveo, to make sure that you're getting the best of the best. Right? If you think about all of the content that you have, not all of the content will be relevant, not all of the content will be high quality, but if you have a relevance engine that can narrow down the most relevant content that can then provide the best answer across any of those apps, and interfaces that you leverage within your digital experience. So when we think about, you know, what kind of outcomes can achieve with total experience AI. You can see some of these are a range of improvements that we've seen with our customers across multiple different use cases. But at the end of the day, if you're thinking about the the total experience, which includes both the customer experience and the employee experience, We know that that creates a more that's more effective for them to find what they need, you know, wherever it is that they are in their journey. And generative answering should not be any different. Right? It should be something that is threaded into the entire experience. So I wanna share some final thoughts. And, you know, again, the first one I've I've repeated a couple of times Generative answering is not a bolt on. It should be threaded into your entire digital journey. Search and content are at the core of generative answering. That's really how you get the best answers. The best experience is connected and effective. And what I mean by that is you know, obviously connected, I think, makes sense, but effective means you're you're getting what you need. Right? And and we saw that with a chat box, it it wasn't effective and part of that is because it wasn't connected. Then the better the overall experience, the more likely you are to meet customer expectations and reach those desired outcomes. So many times I've talked to, you know, customers and potential customers about what they're trying to do with generative AI And oftentimes I hear I'm not sure. I just I'm told that I need to do something with it. So remember that you need to focus on those outcomes and the customer expectations before just implementing the technology. And then I wanna share a couple of resources with you here, which you'll get, when you get the slides. So at this point, I would like pass it over to Brian to, do the demo. Thanks, Bonnie. So let me share my screen here and So what I wanna talk about here, is of course, I wanna show you a couple of examples of our generative AI solution Now I'm gonna be using our partner community. As Bonnie had mentioned, we have rolled this out live on our documentation site. Embedded directly within our own product where we use our IPX. And this is a location that already is being used by our customers. We're already providing a lot of the good information, from just a pure AI recommendation type standpoint. So these are not disconnected silos. If you don't have good search on the front end, you're not gonna have really good relevant generative answers on the back end. Now a couple of things I always like to point out is not every interaction can result in a generative solution or generative answer. So but you're seeing here it's just kind of our regular machine learning recommendations are being returned for a query for a query. You're saying just I just did a query on service cloud. You'll see some recommendation and information being displayed along the top now. This kind of can account for about seventy or eighty percent a lot of times of the queries that are being generated on a support site. Now, however, I'll do a query here showing How does cabello determine relevance? This is an example of where we're taking into account content and going ahead and looking at it from a relevancy standpoint and feeding that into the large language model. They're already doing the heavy lifting on the front end, if you will. So Kaveo has been doing relevancy for many years, and we're applying that type of an approach to actually feed the information into the LLM to create a syntactically correct answer. Now as I mentioned earlier, we're not running this in a in a vacuum. So if I ask a query, for example, you know, maybe I'm looking at you know, how does Kaveo leverage permissions. So here I'm getting a really good general answer, but it's fairly generic. It's asking about just leveraging permissions in general. What if I want to refine that information? What if I'm asking it and wanna interact with faceted navigation. So let's say that I've selected, I'm interested in how Cabail leverages permissions within Salesforce. So now that I've selected one of my refinement elements, mainly the faceted navigation, the entire answer is regenerated. Changed for me. I can do this with any of the elements here if I'm asking about Cycle or something like that. I could also include this within the query body itself. So if I'm asking how does Gaveo leverage permissions within Salesforce, it will also generate that type of an of an answer for me as well. We're not operating just simply on what is being entered into the query. A good example of that as well is let's say that somebody has a profile. So we can take into account context. This is also how we're leveraging permissions and entitlements because one of the key worries that we have or I've seen any way out the industry with people just deploying their own generative answering solution, is they'll take all their documents, put them out there, and you end up with security risk. You end up with trust issues and hallucinations and things of that nature. Since Kavea was doing kind of the front end work feeding that into the LLM, you don't have to worry about a lockdown. There's also a benefit here We ask about something around comparisons. So it's like let's say that I do a query, I wanna compare the feature by feature between a site map and a web connector. So this information can exist and does exist across multiple pieces of content. We wanna show our citations where we're pulling this information from. Also show some additional information that a user can follow-up with and find you know, maybe I wanna take a look at some documents that exist that are gonna be referenceable for me. So when we're taking all of this into account, looking across a lot of information within the unified index, this is where it's really important because we're taking that secured information applying that, you know, combination of lexicon semantic search and feeding that into the LM to create a very trusted and and complete generative answer. Now what happens if sometimes we have a question that may be coming in multiple parts. So here, I'm gonna ask a question explain in detail the difference between Cabeo headless and atomic, and when I may want to use it. So now I'm looking for not only an answer about two different types of features within the Cabela platform, but I'm also asking an opinion if you will. You know, when is the part where should I use this? So what we're showing here is, of course, our citations or references for where we're getting the content. We're giving a great example here of these two type of deployment processes. And then a summarization of well, you know, if you're really wanting to deploy very quickly, you may wanna go with our atomic web components. However, if you wanna have a lot more control over the front end, they wanna go with atomic. So sometimes there isn't a direct answer. The answer is that like depends. And we'll build and generate that for you. This is the only way you can do that when you're looking across multiple content repositories and various different types of information to be displayed here. Now along those lines, let's say that I'm a developer. And I wanna learn the steps to, you know, learn about today with Tom. So now what we're doing is we're again looking at a step by step process. These exist in multiple elements of the content. You can see a lot of the references here. But I'm putting into the query itself, something that could be contained within my profile. You know, I'm a developer or maybe I'm a marker or maybe I'm searching for something I and I have just something in my profile that guides and directs the type of information we're pulling back to actually refine that information and display it in a highly relevant manner. K? Now I mentioned earlier the idea around security and entitlements. So in this scenario, let's say that I ask a question about how do I create a partner organization? Now here it's showing me a generative answer that just really doesn't have the information. Of course, type of a message can kinda prompt the user to say, well, I need some more information. You may not, you know, reformulate what you're asking about. But in this case, I'm not entitled to see that type of information. So what needs to happen here is I need to authenticate. Right now I'm not even logged in. So if I authenticate first, and now I'm an authorized user to ask that type of a question, then the answer that's going to be generated is going to have additional information, if you will. So when I'm asking now that I'm authenticated, I'm logged into the form, how do I create a partner organization? I get a generative answer back. So the difference here, of course, is as an anonymous user, I don't have access to the full content index. We're enforcing those entitlements before they even go to the LMM. And this is really a key point. Along with the personalization and the relevancy, there comes a matter of security. We only want to allow people to see content. And since Kaveo does that natively, we're only pulling back that content that that user has the rights to see, that's being deployed directly within the search result, as well as being displayed within the general answer process. This is how we do get that trusted and relevant types of information. Okay. So now I'd like to close out and, turn it back over to Bonnie. Thanks so much, Brian. Alright, Georgina. Do we have any questions from the audience. Hi. Thank you very much both of you. That was, much appreciated, and it's great to see in-depth on your solution there as well. So, yeah, as Bonnie says, if there's any questions that you would like answered in particular, we do still have a tiny bit of time. Three minutes So, but we do have a lot of questions that have come through more generally about, you know, the nature of generative AI in CX. I'd love to ask you both at seeing what you have presented where you feel that generative AI is going to be the most successful in the CX industry what are we likely to see, coming forward first over the next year, say? What do you both think? Yeah. No. That's a great question. And and, you know, I've definitely seen multiple use cases where people want to average generative AI and generative answering in particular. And I think what we'll see is we'll see some implementations more on the workplace side, some more in internal facing just because, you know, when you think about the risks that come with generative answering, people are less, you know, more risk adverse to have something that may impact customers that may not, you know, be as effective, for example. And so what I'm seeing is, you know, people are seeing the work place use case more as as a starting ground. However, there are more innovative companies that, you know, maybe are more leaders in in tech and an innovation. And, you know, we're seeing them actually implement live with their customers. So an example of that would be one of our customers zero they've they've implemented our solution customer facing, and they've already started seeing some results, which is really great. So, you know, I think it's say people are feeling safer in the workplace side, but it's definitely, you know, I'm seeing it on the on the customer facing side as well. Fantastic. Thank you. Actually, we just had a question come through, I think, probably more for Brian, but I think we should, maybe put it to you, Brian. Is is the the content that you were referring to static or live internal content that you're drawing from currently. And how long does it take to update if the data is refreshed? Sure. That was live content. It is direct on our partners dot cabello dot com site. You can also see the same type of functionality on our documentation site. So if you go to docs dot kabay dot com, that's all live. Now how long does it take for that content to be refreshed? That is utilizing our core Caveo technology. So when we go out and we do a rescan of content, we're doing the incremental update. Let's say that there's new content that's being authored. Once that refresh takes place, that content is immediately available. Since we're maintaining the vectorization within our unified index, then it's updated as soon as content is added. Perfect. And one final question, which is, is the generative aspect limited by needing to provide citations? No. What we're doing there to provide the citations is we're actually showing what content is being passed to the LMM. In other words, what elements that contain the chunks of data we're feeding into that. It's something that we're putting in the presentation layer that we're building out, but you can modify that that layer any way you'd like. Excellent. Well, it's it's certainly very dynamic piece of kit there. So thank you very much indeed for showcasing that, both of you.