Hi, everybody. Thank you all so much for joining us for today's webinar, four ways to innovate your digital workplace experience with AI. My name is Patricia Patsy Liang, and I do content and platform marketing here at Coveo. I'm so excited to be joined by Stan Schroeder, an incredible solution engineer here at our company. He'll show you how Coveo's AI platform powers personalized digital employee experiences, accelerating worker efficiency and your enterprise's impact. To start off, let's do a little bit of housekeeping. So today, everyone is in listen only mode. So if you'd like to ask us some questions, please go to the q and a portion of your screen, and then we'll answer your questions at the end of today's presentation. Today's presentation is also being recorded as well, so please do expect a recording of our presentation within your in your email inbox within the next twenty four hours. With that out of the way, welcome to our live workplace demo webinar, four ways to innovate your digital workplace experience with AI. So here's a brief, here's a brief glimpse at today's agenda. We'd like to start off by letting you know that at Coveo, we practice what we preach. We use Coveo's AI to power Coveo at Coveo, our company's very own employee intranet. As someone who does creative work for many different teams at our company, its personalized recommendations make it so much easier for me to find the content I need to do my work. It unifies our company's knowledge and presents it in one easy to access interface. I don't have to search through Jira or Google Drive or SharePoint to find what I need. It's all here in one centralized hub. That's why it's not surprising that AI is more popular in the workplace than ever before. Here's just a few market trends for this year. As we can see, four out of five CEOs are increasing their digital technology investments to counter current economic pressures like inflation or scarce talent. Forty six percent of business and technology decision makers are seeking out partners to implement AI critical to their businesses, and seventy four percent of companies that implement AI as part of their tech stack see a positive impact across their organizations. And that makes a lot of sense because AI has the potential to help us get more done with less resources than ever before, helping us save time and energy so we can be even more productive. Now let's move on to employee expectations. Here's the digital workplace experience most employees expect. They expect to be able to access company knowledge easily from wherever they're searching from. However, most of us know that the digital workplace experience really isn't that linear. It can be really hard to find knowledge, especially when it's kept in multiple disparate silos. And that makes a lot of sense because research shows that workers spend nearly half of their day searching for the knowledge they need to do their jobs effectively. That's a huge loss of time, energy, and resources. To make matters worse, all of this extra work makes employees feel very burnt out and more likely to quit. Well, that's where Coveo's powerful AI platform comes in. Today, we're going to show you four ways to innovate your digital workplace experience with AI to help you see improved digital experiences, lower costs, and higher revenues. We'll be exploring Coveo's AI for the workplace in four different use cases, including employee intranet and self-service portals, unified knowledge management, personalization and recommendations, and rich analytics. Take it away, Stan. Great. Thank you, Patricia. So I wanted to start today's session with a little word puzzle. I heard this, this weekend on, National Public Radio, on NPR's weekend edition, and I just thought it was so great. I wanted to share it with you. So if anybody has already heard this or they know the answer, don't say anything yet. And don't anybody chat in any answers yet. Let's wait till the end of the session today to see if anyone can figure it out. But here's the puzzle. Name something scary in two words. Five of the letters are vowels, which are all the same letter, and the consonants are all Roman numerals. Now extra points if you can give us the sum of the Roman numerals in the two words. So five letters are all vowels, and the the all the consonants are Roman numerals. Alright. So let me get started here. I'm going to, share my desktop so you can see the demonstration. And, let's see if I can get this to show up. Okay? Alright. And, Patricia, can you see my desktop? I can see it perfectly, Stan. Fantastic. Thanks very much. Okay. So the idea behind the Coveo of Workhub is twofold. We wanna increase, your employees' efficiency by making it easy for them to find what they're looking for across the enterprise. The second part of that is increase their job satisfaction by reducing the effort for them to get the resources that they need to do their jobs. Well, how do we do this? First, by streamlining the digital employee experience with an intelligent back bone in the form of AI powered search. And second, by approaching the employee's search experience using some of the same techniques that we use for our customers. Now we've all heard of the consumerization of IT, which means we all now expect the same level of quality in our business lives that we get in our consumer lives. And that includes important elements such as personalization, recommendations, search quality, and features. I work with clients every day, and when they talk to us about these kinds of projects in the context of enterprise search for employees, The most frequent way that they describe this to me is that their employees want a Google like experience, and that and that's a quote. And and what I mean by that is I get that quote same quote multiple times. At Coveo, we wanna help companies drive resiliency, foster connectivity, and improve the digital workplace experience for their employees by making it easier for employees to connect with just the right people, adjust the right information, adjust the right time. So let's take a look at what we mean by giving employees the kind of search tools that we give our customers. And to do that, what I'm gonna do is I'm gonna look at a couple of our customers' websites. So for the first example, this is Bunnings. So Bunnings is, the the Home Depot of Australia and and the New Zealand area, and they power their ecommerce site with Coveo. Now they use a lot of different parts of Coveo, but the capability that I wanna bring to your attention here specifically is how they take advantage of our powerful and highly configurable search bar. So I'm gonna click in here to the search bar. And before I start typing anything, first thing you notice are the popular searches, the popular products right now, and, some content recommendations here on the far right. As I start to type in my query, let's say I'm looking for cordless drills, you can see I'm getting intelligent query suggestions followed by inline search results. So what does this look like from the perspective of an employee work hub? So as I go back here to my employee work hub example, I click into the search bar. Well, what you see here are, recent searches. As I start to type in a question, you see, instead of the feature results, I'm actually getting inline search results. So very, very similar, to what we've done for, for for Bunning. Now for my second example, I'm gonna go to another one of our customers, Salesforce. So, Salesforce is a long time Coveo customer. They use Coveo in lots of different parts of their business. And and in this case, this is the, AppExchange, and this is a Coveo search powered, website. But I actually wanted to bring your attention to these recommendations here inspired by your searches and then the sponsored solutions below that, etcetera, the customer stories. This information is actually driven by Coveo's recommendation engine. So, built into Coveo, is a powerful, general purpose recommendation engine that allows you to, provide personalized suggestions to your customers. These are personalized, by the fact that I'm logged in and it's associated with the the user and the org that I'm associated with, as this particular user. Now what does this look like from the perspective of an employee workplace? So for that, let me go back to my example here. We call this the, Barca Crew Workhub. So Barca is our, fictitious demonstration company. They are a manufacturer of of of boats and sailboats, motorboats, marine engines, and related equipment and merchandise. Our story is focused on one of their employees, a designing a design engineer named Steve. Steve's gonna use the Barca crew employee work hub to answer some questions that he's got, and this work hub is powered by Coveo's AI driven engine. And it's personalized to Steve to maximize his efficiency and relevancy. So, Steve, as you can see here, is a design engineer. He's based out of the US, and, his specialty is design. Now, if I actually change roles for just a sec, and let me use as a different example, Regina, you can see here when I log in as Regina, the recommendations that she gets are really focused around, the fact that she's an HR specialist, and and her specialty is payroll. And she she's also located in the US. And you can see here, the the the content that's coming up for her is based on what other HR professionals have indicated as as useful content. And it's stuff this, that, sort of focused around, things like workplace transformation, HR, payroll topics, things like that. If we go back to Steve, you can see very non HR professional stuff. It's much more focused around things like, organizational improvements, remote work, values, things like that. Now but beyond just the content here, and there's different kinds of content that's being, presented, in these two recommendations. The content on the top here is stuff that we've indexed that is, part of our organization. The videos, these are actually external videos coming out of, YouTube. So there's both internal and external content that's being represented here in the recommendations. Beyond content, though, on the right hand side, notice these two other sets of recommendations, the peer influencers and topic influencers. So these people recommendations, aren't content in the sort of the traditional way. ARC is a large organization with thousands of employees representing multiple brands and product lines across many states and multiple countries. Beyond just doing people search, what about facilitating networking within the company? Well, this first list is recommending influential peers to to Steve. Other designers and engineers and engineering managers within the organization. Maybe as a boat designer, he might wanna reach out to an engineer who writes specs for motors to ask questions about motor mounts and stress factors. The second list is the people recommendations of people that have the same specialty as him. So maybe he's interested in exploring using some new design or simulation tools, and he wants to, talk to other designers and and find out what they're using. Coveo's recommendations are built by the integrated AI recommendation engine and can be deployed from multiple use cases like you see here, or like by the the Salesforce example that I showed you earlier. But Coveo isn't just a recommendation engine. People still need great search, and they want that Google like experience. So let's try that out. So Steve has heard of a new internal job program in Barca called Barca Gigs. A gig is where you can temporarily be assigned to another department to try out a role that you're interested in. Steve's got an interest and a background in data engineering, so he's been wondering about the possibility of trying that out. So first, he wants to know exactly what is a gig. So let's go ahead and and ask the question, what is a gig? Now, what you'll notice here is Coveo smart snippets feature giving him an answer to his direct question. Alright. So great. He's got an idea of what a gig is. So let's say he's got a follow-up question, and he says, what skills do I want to develop in my gig? And we get a, a direct answer from Coveo smart snippets feature. So smart snippets is an AI large language model that works off of HTML content and discovers question answer types of combinations in the content and uses that information to help the user by providing a direct answer as a way to help them save time. Now, maybe he's got another basic question. He wants to let's say he's like me. And this weekend, he wants to finish his his taxes, and so he's gonna need his w two. So let's go ahead and let's ask a question. How do I view? And we'll take my or your suggestion here. How do I view my w two on, form online? Now you'll notice that the system is trying to help him, in multiple ways here. First, you you can see the system has opened a pop up box to let him know that he can talk to an HR or payroll specialist right away through a Slack, Huddl, and providing that link to that Huddl. However, if he, if we don't want to sort of interrupt the the the user with the pop up box, there's another type of, trigger rule that we can run, and we call this a notification trigger. So the first, trigger with the pop up box, we call that an execute trigger. This up here, this little message up here, this is called a notification trigger, and you can put any HTML inside this as well. The idea is that this is a nice little reminder. If at any point they need to, you know, click on, the the link, they can access a Slack huddle and talk directly to an HR specialist. And then the third way is the system is trying to help him is, through yet another smart answer. So the smart answer, the smart snippet is giving him the answer to his question. The AI generated smart answer is based on what it found in the index. Now let's, talk about another scenario. Let's say he's, since he's got this background in data engineering, maybe Steve wants to, go back to school and and further his education, but he's got some questions about tuition, assistant at Barca. Now, the first thing that you'll notice is as I'm typing in the question, the query suggest AI engine is making, suggestions to help him find an answer. The this is a purpose built ML model, that comes out of the box, and and this model works on previous successful user interactions. So it's not just questions that have happened a lot or or queries that people have put in a lot. These are specifically around the idea that people, like Steve with his kind of profile, have asked these questions and have successfully found an answer. And what this does is this help users form better queries. It helps users with spelling, help people overall improve their likelihood of success. I actually use this feature in Google quite a bit, just because I'm a poor speller. So if nothing else, that that's a really, really useful tool. Well, Coveo got that built in, as one of our purpose built, AI models. You can also see the, other ways that the the system is helping, Steve. First, in the smart snippets, it's making suggestions on other queries that people have, asked or performed, successfully. And then but I what I really wanted to draw your attention to were these featured and recommended flags. So a featured item featured flag indicates an item that is showing up because a manually derived relevance rule was put in place. This means that somebody configured a rule that said, when this type of question is asked, we wanna show this document or these documents at the top of the list. The recommended flags indicate items where the AI engine has figured out that these are also useful art, or automated relevance tuning model is automatically showing Steve what others have found useful when they're asking, this kind of question. And it's these are happening right alongside what the the search administrators have put at the top of the list with the featured items. But what's important about this is that they work together. There is no need for someone to try and figure out or manage the AI engine. The two work together seamlessly and automatically. The other thing I'll point out here is as I scroll up and down on the right hand side, you'll notice that there are filters and facets there. There's yet another AI model at work on those that we call dynamic navigation or DNE. I'll talk about that a little bit later on in in our presentation. But in general, what I wanted to, point out here was this simple query, there's no less than four different purpose built out of the box AI models at work, helping Steve find exactly what he's looking for and answering his questions in a personalized and highly relevant way across any content that we've indexed in the enterprise. And we talk about out of the box. These are quite literally out of the box. The configuration for these is is takes minutes. It it's like, literally just point and click to get these, models up and running. Now let me go to another scenario. This time, we're gonna talk about, personalized search result. So, to do this, let me introduce another character in my, my little example. We're going to talk about Jesse here. Jesse, is a manager. He's located in Canada, and his focus, or specialty is on safety and, safety engineering inside of, manufacturing plants. And so that's Jesse. But, when you look at his recommendations, you'll notice his recommendations are, again, different from Steve, different from Regina. Likewise, the peer influencers and topic influencers are also specific to, him and his profile. But, really, what we wanna look at here is, let's say he's got a question about, viewing his, viewing his pay stubs, let's say. Now you'll notice because he's Canadian, the the content that's coming up for him is based on the fact that he's he's in that region. Whereas if we go back and, perform the exact same query as Steve, who's located in the US. You can see the content that comes up for him, is focused around the fact that, he's a US based employee. That's what we mean when we talk about, personalized results. Now, these can be driven by manual rules, filtering rules, or boosting rules, or, the AI engine personalizing experience, or even both working together. So really, what we're, what we're talking about here is the ability for you to, personalize and and make the content, the search results relevant to what, each one of your employees is doing. Let me do another example here. I'm gonna go in this time as Regina. And Regina, as an HR specialist, I'm gonna do a very, very generic, query here. I'm gonna say she's searching for training. And what I really wanted to show you here was how asking a the the exact same question as, Regina, Steve, very, very different results. So you can see here, Regina's, content is sort of slanted towards, like, legal and regulatory topics around training. If we go back in as Steve, who's a a non HR person, and when he searches for training content, it's much more tuned to, you know, what other, employees that are non professionals, might be interested inside the organization. Now, while I'm here, what I also want to talk about was what we mean when we talk about the universal index in Coveo. Coveo can index basically anything in the enterprise, whether it's content that's in the cloud, on premises behind the firewall in a data center, structured or unstructured content, packaged applications or custom built applications or databases. And when you you when I, actually go inside the the the different tabs here, having done this very general query where that I know has content across all these different tabs, you'll notice here as I click into the tabs on the right hand side how the filters and facets change. They're contextually aware of what it is that I'm looking looking at. You can see here when I'm looking at, at the Slack, tab, I've got, Slack channels, Slack user. When I click on the my barca knowledge base tab, I've got, you know, things like article and region. When I click into the well-being tab, I've got, keywords and tags. Likewise, when I click into the crew members, I've got titles, location information, office location, departments, etcetera. So these these the filters and facets are are are contextually aware. Basically, any metadata that exists in any of the content that we're looking at can be used by the index, and we can use that in multiple ways. It could be used for, driving relevance. It can be used like you've seen me do here with, filters and faceting, or can be used for, you know, purely viewable data, and it's that flexible. And, we have the ability to normalize, the facets across disparate sources as I mentioned. We also have a prebuilt out of the box ML model directed exactly at the, at the facets. So we call this dynamic navigation. And what DNE does is it actually looks at the usage of the facets in, associated with specific queries, and it'll reorder the categories of facets. And within the categories, it'll actually reorder the links within those categories. The clearest example I can give you on why this is important, let's say if you were shopping for a bicycle, and then the the facet color would be very, very important, so you'd want that to show up at the top of the list. But if you were shopping for a bicycle chain, well, color doesn't matter at all, but the facets that might matter in that case would be the length of the chain, maybe the material it's made out of, or the weight of the chain. And so you would want those, facets to show up at the top of the list. Well, that's exactly what DNE does for you. And it can get so smart that it it can actually preselect the facet for you based on what other people have used as far as the associated with a specific query. Lastly, I wanted to leave you with the understanding that, out of the box, Coveo is, providing, out of the box analytics like all of the Coveo products. Coveo automatically captures every interaction in our session database. The information is used by both our AI engine as well as, for our out of the box reporting and usage analytics, like you see here. The the Coveo a AI engine learns how to serve the best content with the least effort across multiple use cases and touch port touch points, and then you can report on all of that usage. You You can report on, specific content usage. You can report on the content gaps, etcetera. I don't really have a whole lot, enough time in today's, session to do a a a lot of demonstration around the, the reporting and analytics. But all of the data that you see here is available, directly in Coveo, or you can also download this information or access it through, the reader account into the, into the Snowflake database, directly. So there's all kinds of ways to access this data and and, use it within your organization. So that's it. That's my my demo for today. What I'm gonna do is, let's switch back to, Patricia's, fly decks, and let's, review the the puzzle that we started with. Let me stop sharing here. And Yeah. Patricia, can you bring up that slide? Yes. Really great presentation, Stan, by the way. Thank you. Can you see the slide? Not yet on my end. Oh, interesting. Here. Let me run around a bit. There you go. Alright. And Alright. Oh, there's the the answer right there. So so the the the puzzle was name something scary in two words, five of the letters or vowels, which are all the same, and the consonants of Roman numerals. What scary thing is this? I don't know if anybody, figured it out and chatted anything in, any answers in. But the answer is voodoo doll. I thought that this was such a great puzzle. I I I wanted to start with it. Now and when I figured out the, the answer to the sum problem, I came up with eleven hundred and five, which equals MCV in Roman numeral language. So hope you guys enjoyed that. Do you, wanna, wrap things up, Patricia? Do we have any q and a's? Yes. We'll keep going. Very cool. Two words that are scary, manual tuning, bad search, my in laws. Really great and really cool. Alright. So hi, I don't know if, Patricia's, Internet froze or not. But You may have. Can you hear me okay, Stan? I can. Yeah. Okay. Well, I will I will help to wrap things up. She did freeze, actually. So Oh, that's funny. Yeah. Thank you again everyone for attending. I'm gonna see if I can share my screen because we did just have a few slides we wanted to share with you before we end here. And we'll start from here. Can you see that okay? No. I'm I still see the in conclusion slot. Oh, there we go. Okay. It's just switched over. Awesome. In conclusion, at the heart of a great digital workplace experience is a flexible, composable tech stack. We hope you got a glimpse of the power of an AI platform, and we'd love to show you more. So with that, we'll share with you a couple of upcoming, events or ways that you can get in touch. So we do have our our three sixty, new and Coveo sessions coming up over the next couple of weeks. Our session on workplace is on March thirtieth at eleven AM eastern time. But we also have an event on March twenty third, which, will cover some exciting new product updates as well. In addition to that and welcome back, Patricia. We did want to offer you two links. Again, we'll share those here. The opportunity to book a demo, and if you're an existing customer, you can always reach out to myself, Stan, or Patricia through your existing CFM. We'll go into more detail on, how you can create a, what I would call, a a world class employee experience. And if you're more on the techie side, you'd like to just jump in and get your hands dirty, we have a free trial that will share some links with you as well. Yeah. It's really fun to experiment with. And take it from me, Stan, and Juanita, you won't be disappointed. Yeah. So do you wanna cover some quick q and a? Yeah. Let's do some quick q and a. So first in the chat, I can see, does my content and text does my content and taxonomy need to be optimized before I get started? Oh oh, that's actually a very great question, and it's a common question so the answer is it doesn't have to be. We will use whatever existing structure you might have, in your content. So, if your content is organized and tagged into a taxonomy, or an ontology or structured in some way, we can absolutely use that. And we can use that information, as I said, for driving relevance. We can use it for, you know, filters and faceting and and and segmenting content, etcetera. But it's not a requirement, and a lot of organizations feel like they gotta they have to do that first before they deploy, Coveo or or any search. And the answer is really you you don't have to. And if you wanna add that on later, Coveo can, still make use of it, subsequently. But Coveo has some specific capabilities that I would say mean that you don't have to do that upfront. One of the things we do, is we can autumn we automatically, assign concepts to content. So we have a built in, text analytics capability that assigns, it looks for interesting words and important terms in the content. It does some basic, entity extraction, and it'll automatically assign that data, to the content as we're indexing it. And then as I said, you know, if you add that subsequently, we'll we can, include that as part of the index. More importantly, the reason why I suggest to to my, customers not to wait is because what you lose then what's the economic term? Opportunity lost is this whole bunch of learning that you lose in terms of the Coveo engine. As people start doing their searching and clicking and that sort of thing, that's not dependent on those taxonomies. And, you know, if you delay the project by six months, that's six months of AI learning that you don't get the the advantage of. And and in my way of thinking, that just accumulating that data, even if you don't use it to drive the relevance in the you know, right away, that's data that the AI engine really, really can use. And more data to AI is better better relevancy, better accuracy, etcetera. Very interesting. The question. Yeah. I think we have time for one more question. Let me see. K. Does Coveo use a list of prebuilt synonyms? So no. We actually don't. However, there's, there's just a plethora of, you know, synonyms out there on the web that you can download, and we can take those and bulk upload them into the system. But what we have found is that generic English language synonyms, for example, are really not that useful. Our best practice on this area is to say, make the synonyms specific to your organization because everybody every organization has sort of their different lingo. And and, generically, just throwing a bunch of, you know, untested synonyms, can have unintended side effects. And so the the best practice is put in the the synonyms, the queries, the acronyms that are specific to your organization. Start there. And then, through the the sort of a normal user acceptance testing phase of an implementation, we can enhance, that list by putting in, you know, terminology, stop words, and and and other terminology that may be useful. And we have actually multiple ways to define a synonym. Don't have enough time in today's session to go into the detail there. But let's just say that it's not doesn't have to be as simple as, you know, synonyms and acronym expansions. That's so cool. Yeah. If you wanna learn more, book a demo with us. But that's really cool. Not to mention the fact too that, like, Coveo works with hundreds of languages too. So the synonyms is really important. So I think that's all the time we had today for our presentation. Thank you so much for joining us for our webinar, four ways to innovate your digital workplace experience with AI. Thank you so much. Have a great day. Thanks, Juanita. Thanks, Dan. Bye.