RDA

Hi, everyone. This is Gina Baumgartner from RDA, and I am your host today for our for our second Cycor summit series, and I'm super excited to be joined by Ali And Ali, I I tried to pronounce your your full name, but I haven't mastered it yet. So I'm just gonna say Ali. Ali, you are, the product manager for, for Coveo for websites. Could you tell us a little bit about yourself, and, you know, about Coveo and what is, I'm I'm super intrigued by your title. Could you tell us about, you know, what does product manager for websites mean for, or Coveo for websites mean? Sure. Well, I've been quick though, to start for about five years now. And before that background is in finance unrelated completely, but I had a computer since I was three years old. So I've always been geeking out and, spending way too much time on the computer. I do my work day and then I spend some more time with them. Right? Yeah. That's how you helped me debug my camera and and video and audio sync issues. Right? That's exactly it peaking out on stuff like this fancy camera set up right here. Yeah. When you don't have to the best to do the thing. So, that's one thing. For the role of product manager, it varies between companies. It varies also internally but usually you act as a conduit between, the external of the market, internal between teams, coordinate, coordinating efforts, prioritizing what is most important. So taking decisions to ensure that we deliver the best product possible, it can really vary are more technical or less technical product managers, more coming from a business side or coming from a programming background. So it it really depends And as far as Cabail goes, well, we are a, platform that focuses on relevance at every interaction. And I'm gonna go a bit further into that later if we have the chance, and I think we will. Awesome. Well, our topic for today is called AI Search. So I'd love to hear from you in terms of What is AI search? I think the most important distinction to make is, the difference between static search and dynamic search, so AI powered search, where if you start with something like a solar, for example, and you don't have a data science team. You don't integrate anything to continually improve your results. Well, you're gonna have something static that in the first year or in the fifth year, your results are most likely gonna be the same except for force adding things. When you have AI powered search, we take into account user interactions on your website. We collect usage analytics data, and we use that data to improve reranking, providing query suggestions, taking into account the intent of the user to determine what is it that they really want. Right? We wanna make every single experience that they have on your site relevant. And when you think about the kind of experiences that people expect today, and it's it's easy to refer to Amazon YouTube Twitch, you know, all the big players in the industry But that's the reality. The reality is that people, are used to this kind of thing. They arrive on a website. They arrive on Amazon. They get query suggestions. They get nice fastening. They get recommendations as well. Right? They have all of those things, and they take it for granted. But it's only when you don't have those things that you realize that something is missing. And AI powered search kind of encompasses all of that. It's this whole experience of taking, what people actually are doing into account and improving results. Yeah. It's pretty cool. It's like you guys are commoditizing what the Amazons and Netflixes of the world are doing to everyone or every website. Exactly. Because, at the end of the day, when you think about building all of this stuff yourself, you start with, you know, solar instance or Azure search or whatever you're doing. And you wanna get to the point where you even start competing with all of these companies that's a lot of work, right? And just getting to barely competing there, you will have invested a ton of time, of effort, of money, And anything else that you want in the future, you're gonna have to put in more effort, invest in a in a linear fashion, and, just keep adding, keep maintaining a more and more complex product. Whereas with a platform such as ours, from the get go, you are competing with the big guys, And then anything else that you can dream of in the future, you have a a huge set of tools at your disposal to build anything that you want. And I'm I'm not trying to sound too much like a salesperson. I feel like sometimes I do. I'm just the only things that I can sell are things that I truly believe in. So I drank the Kool Aid, but I I really do believe in our product. So I'm I'm gonna calm down now, and I'm gonna let you ask more questions about that. Sounds good. Sounds good. No. I love I love your passion. I think that, you know, passionate people, are able to really present it in the light that makes sense to everyone, and and I love your passion. So going into, you know, our our series is called the psychros Summit Series, and obviously your focus is also Coveo four website. Which, Sycor is is a big Coveo partner and, our partner too. And we'd love to talk I'd love to talk more about okay. So what could you do with AI powered search on websites and specifically on psychcor. How would how does how does Cavell for psychcor work? Well, my role initially was So product manager for Kever for Cycor. Our Cycor integration is, has been long running a very mature product. That is basically the best way for our micro customers to harness the power of the conveyor platform itself. It that's where the power really lies at the end of the day. But all of the tools that you have when you install the Care for Cycle package are really essential to get started quickly. So we have our search UI framework called the high framework, giving you all the building blocks you need to create, search pages, add recommendations. So you have, you know, facets, the search box, the results themselves, everything customizable, available as MBC renderings. We have indexing capabilities built in. So you have your crawlers by default that'll take your you know, your master and your web databases, index everything fully customizable. And we support things like multi language, multi site, So everything out of the box is there for you to really harness that. You can add rules in the front end. You can do it on the platform as well. But when it comes to Kavir for websites then in general, we basically wanna make sure to recreate that kind of experience everywhere you go because you're not limited by the fact that, a site core, for example, has a deeper integration than some other CMS or some other, you know, experience platform. You can do everything still. The platform itself can index content from any source And you can use our variety of different UI, elements that we have. So we have our UI framework that's similar to what we have there in sitecore based on our job search framework, but we also have new ones that are headless. So it did come without components. They only handle the logic to communicate with our platform and you can build your own UI layer on top of that. Awesome. Well, without further ado, you know, one of the big portions of the summit series is actually demoing demoing to our audience. So, sh, could you show us what Coveo for cycle actually looks like and what these components that you just talked about, how they work. I'm super excited to see it. Absolutely. Let me see here if I can share the right screen. Alright. Perfect. Yep. I think I think you see that. Right? Yeah. Yeah. I see it. So we do happen to have, some things that RDA worked on. I would have chosen them no matter what because they're great examples. So starting with security brands, one of the things that I really like here is the these guys have complex products or complex product names And, when I look at it, of course, to me, necessarily, it doesn't mean anything. If I hit the search box here, I'm gonna start seeing query suggestions I can type in something like, CPX here. And I'm gonna see stuff like this. Now query suggestions are based on what users actually typed in. They're not based, for example, on a field value that's being displayed, but there's an entire flow of a user typing something, arriving on a search page, and then actually clicking on a result, for example, or refining and then clicking on a result, All of that chain of events has to be successful for query suggestions to start appearing. So we know that when we suggest something, it's going to lead to a successful outcome. So I can click on on this, and it's going to lead me to, god knows what product, but it seems like this is the most relevant one here. So CPX LED flat panel, that seems to be two by four. So that's most likely what the user actually wants, and the the recommendation is here. I know they're on. Okay. That's that's so cool because, like you said, you know, those words mean nothing to us laymen, non lighting, non lighting experts. But Coveo learned that I mean, this is a this is the true power of AI search because Kavell learned that query, right, because people that's what people are searching for. Absolutely. If I click on this now, I'm just gonna reinforce that model based slowly over time. It takes a certain amount of interactions, depending on how many people visit. But, yeah, you also have, on the site here, the product pages themselves, if I go into a product page, this is going to be a Coveo powered product page. So where you have, again, large number of facets that can allow you to sort through all of these options that you have on the complex products. You can see more products. You can change the layout. Really, really nice. Very cool. Something that we're pretty good at, and, so this is a a more recent implementation here. I think on your guys' slide. Yeah. That's right. Law firms. Law firms, I mean, if you are a lawyer at a law firm and you can't find yourself on the website, someone's gonna hear about it. Right? So you really need to have a great people search. And on Covington, on cloud dot com here, we have a combination of people search and a regular result search, in the full search page. When you actually start typing here something that's interesting is that These are not query suggestions. These are actually result suggestions. So, that's something that we actually see more often with law firms. I don't know why it's a common requirement, but it's something that can be done if you wanted to. So you can have results directly. And these are also scoped down. So if I go on to professionals tab, and now I I I hit the same thing again. I'm only gonna get the names here. So it's very customizable in that sense. Navigating further down into the pages. We can also see that this has a Cavell powered aspect to it. So we're pulling information about, Scott, a Anthony here, who is a partner at Compton, and we're displaying related work, other alerts, I believe. So everything that has to do with him and his area of expertise directly on the page through Cavell. So that makes the page more dynamic, and it's gonna evolve over time as things change. Very cool. Yeah. I love this site for sure, and it is a recent recent development with Coveo, but it makes a huge difference to the website. I love it. I don't know why, but the the the law firm websites always have the nice clean pictures. It's it's a it's a really nice experience to have. Yes. Although my favorite one is the the Virginia ABC site. So Boos is easy to relate to. Is that is that a biased light thing and, and lawyers don't appeal to you as much, Ali? I don't have any significant feelings, but no, bias lighting. Look, look at that lighting over there. I I do appreciate that lawyer. I'm indifferent. For for the Virginia ABC site, we do have a great very suggestion. So when you start typing something like Verban, we're gonna have, all sorts of different complex queries as well. If you actually go visit, the page directly here, we're gonna have our recommendation component. So The recommendation component, it can seem here as we're just pulling up, you know, related items, but that's not really the case. What happens is that when I clicked on this item, we sent a page to event with information, from, from, the visitor, to Cabail Cloud, and we gather analytics on the patterns of browsing behavior that the users have. So that's how we start recommending these things. But one thing that you can really see here is if you start typing something, and we actually have it here, I think we use it maybe too much or people actually buy this. It's called Vaca. If you type this as an example, and we go see, for example, this crystal head vodka. What am I going to get as recommendations? Well, I do get vodka, but that's not really what we're focusing on here. What we see is that people who tend to buy crystal head vodka, they're not looking for vodka. They're looking probably for novelty items. Kind of gift where they're doing birthday gift, Christmas gift, and they want something that that's fun. So it could happens to be vodka right here because we're taking that into consideration as well, and there's enough novelty items to do that. But this is where the intent is. Right? We we don't wanna show them grey goose. We don't wanna show them anything else. We're showing them novelty items. Virginia, BC as well have a very nice, product page listing where you can go into different types of of alcohol. And the full page here don't have to apologize my internet. At times, it is not the best. No problem. We can just chat. Well, I know what I know what I'm buying for Halloween this year, but crystal skull vodka. I'm not sure how good it is. Imagine that it's alright. There we go. So if I'm on a bird bourbon product page here, so we have a really nice listing page here, providing everything a really easy way to browse, you know, by names, by sizes, I different types, everything that you need. I I really love these product powered pages where we gather also the analytics data. So you can have that for reporting if you have a product page, super synchronous. Awesome. Yeah. This is great. One last example that I'll show is our Habitat home demo. And the reason I like this one is because it uses our integration with X DB to pull contextual information. So, when I go into the search box here and I search for speaker, what we're gonna have, when we land on a results page is a variety of different results. So, yes, they're all speakers, but we see that we have weather, outdoor speakers, floor standing ones, lower. We have car speakers, bookshelf, ceiling speakers, everything there. And on the site core front, we can see that we have this little menu here. We haven't been profiled as any type of user yet. So if we visit visit a page about setting up a home theater Right? We've set up some, pretty aggressive, profiling here, but will be identified as someone who is into home cinema. Now if I go back and I do the same query, I'm going to get an entirely different set of results. I'm immediately getting home cinema packages, amplifiers, sound bars, things like that. And this isn't even machine learning, right? The only thing that we're doing here is that we're taking into account the fact that now we are identified as, you know, a home cinema enthusiast. We're sending that contextual information And in the back end, this is the Coveo, platform, administration UI, we can add a rule that says if the user is identified as having, you know, home cinema, in their in their profile, we're gonna boost a category, a computer category here of theater. It's something that machine learning will figure out because we're sending the the information But even if you want to do it manually, you have the ability to do so because on the platform side, and unfortunately, I don't have the time to go over everything But the concept here of query pipeline where you can apply these different ranking rules, add things like, you know, the source entries, configure machine learning, do really a bunch of things, do things like add featured results, for example, as well. It's really powerful. You have everything that you need to manage, you know, your reports, your data, your different sources of content that you wanna add So right now, we'd have two site courses only. But if I want, I have access to just a slew if that's the right word. I feel like I'm not pronouncing it correctly. Of connectors. That's the correct typing. Yeah. I think I choked on that one, but we have a variety of connectors available So you can get your content from any number of sources, build your search pages, collect usage analytics data, and create experiences that people identify with that they relate to based on what they use most pretty much every single day. Yeah. That's great. That's great. I love the, you know, I I do a lot of personalization work with SaKor and setting up profiling and goals. So the ability to use the power of Cycor along with the power of Coveo is a very compelling argument you know, it was pretty cool to see the results immediately change on the on the habitat page in terms of, looking at just the focus home theater stuff knowing that that's what you're interested in. Absolutely. And things like query suggestions will also shift over time. So everything will shift over time to, take into account that context, everything we know about the user, everything that you give us about them, if it's statistically relevant, we'll take it into account. So the model figures it out all on its own. Very cool. Very cool. Well, sorry. Did you have more to show us Ali? I I did not. No. Okay. Perfect. Awesome. That sounds great. So, Ali, we, you know, you showed us some really cool features of of Coveo for Saikor. When is the right, you know, for our audience watching us today if they're trying to make the decision in terms of is Coveo for Cycle for them? When is the right time to use Kaveo for Cycle? And who should use Again, here, not to sound like a salesperson, but the question should rather be when is not the right time to have something like If you don't have search on your site, you probably should. There are some rare exceptions. And if you do have search, then you shouldn't do it halfway. You need to have something that is competitive that really serves to drive the whole experience because the search box at the top of your site is the way that the customer talks to you. Right? They tell you literally what they want. And if you have the ability to harness that, harness what they're doing, and, use something like Coveo to use our platform to really drive those experiences, provide recommendations, then it's always the right use case. And it can go from a very small site to huge enterprise implementations, everything is covered. So we have a variety of different plans to to suit the different needs, so yeah, how much time do you want to put into building your own solar implementation? Yeah. I think I think that time would be better spent actually thinking about building the experience. Yeah. And over time, you know, you're spending a lot more unable to keep it up to date with the changing technologies, whereas if you go with a product like Avaya, where it's SaaS, you're constantly pushing out updates, keeping it up to date with the newest technology. Over time, you're definitely saving a lot of money by going with Coveo for Sycor. I mean, we have more PhDs than our competitors have employees. So Are you one of them, Ali? Are you I'm not. No. Okay. I have I have a pretty big head, but it's a regular sized brain texture. Alright. Sounds good. Well, thank you so much, Ali, for, for sharing your knowledge and and your time with us today. A couple of things before we leave, here is Coveo blog, for, you know, for anyone who wants to look at, more information around Coveo for Cycor. And, we are offering everyone who is joining us today a free art of the possible workshop, which answers five questions. So what we will do you know, look at your your industry's stats and show you what your customers want, where your competitors stand, and then look at an art of the possible that's focused on AI search. And, Coveo, we'll do a one on one customized demos. If you wanna Ali went through a few features really quickly in ten minutes, there's a lot more details behind that, and we can customize that demo for you. In terms of what Coveo could be for your company and your website. And then finally, our next Sycor summit series, our third one, will be on content hub and featured on June twenty fourth. So we hope to see you all there. Thanks, and have a great day. Thank you. Have a great day.