All right, I think we are live, Mr. John.
Great. Hey, Kyle, and hi, everyone. John Murcott here. You will recognize right away that this is not the deep, beautiful radio voice from Jonathan Meyer, who is out today. But I am very excited to have Kyle here, so let me just share my screen.
Very excited to be here as well, and welcome, everybody. Happy to have you.
Yeah, I'm definitely excited to talk about our new release, Athena, follow-up to Zeus from our last presentation. I run product and strategy here at HawkSearch for people who have joined before. Kyle, could you do an intro yourself?
Yeah, absolutely. So I've recently taken over the position of Customer Success, so super excited to change the career path just a tiny bit. Before, I was actually heading up the Professional Services Department. And then before that, I actually got to spend some quality time with John Murcott on the product and strategy side of the fence. So it's kind of come full circle, but super excited to be spending more time with customers and ensuring their growth and success, of course.
I appreciate, Kyle, you bring a lot of hands-on experience to this. You know this message, just to spend a quick second, because I always try to figure out when we're working on the roadmap and coming out with new releases, is it always addressing or moving ahead our core goal? That is relevancy, because the more relevant the results are in your search, the more conversions you're going to have. That hopefully drives all of our roadmap, and we'll confirm that as we go through this. It's especially with B2B. Most B2B users use search on a site. When they use that search, we want the results to be relevant. The more relevant the results are, the more likely they're going to click on those results and drive a conversion.
So that's what's driving all of our roadmap and where we're trying to go with the world. But what we found, at least previously, before AI came into the picture, is that we were always, and the market in general, was trying to squeeze all of that relevancy from keyword search. It's like, how can we add more and more capability to keyword search to get more and more relevant results? But now, in the world of AI and vector databases, et cetera, we now have two huge new opportunities to drive more relevancy. And those are namely Concept Search and Image Search, features we've talked about before, but what we'll continue to talk about, and we're adding to the roadmap here. Kyle, any thoughts on where we're going with relevancy and what you see in the market?
Well, the only thing I wanted to maybe quickly call out, I know that yourself and Jeremy, another member of our team, had gone to a conference. I think it was mostly around it was for distributors, mostly around AI and how it influences the marketplace. But I think one of the key takeaways that Jeremy had mentioned to me was image search resonates. And when speaking to distributors, they immediately see the value in image search and are key into it and want to discuss it. So I think that's pretty cool. And I think also that at one of the presentations, one of the most important things, the first thing that was brought up in that presentation was search and how distributors are using search to help drive their marketplace and their growth.
Totally. Absolutely. All right, so let's just spend one quick minute then on the roadmap itself. Where is this all going? Where have we been? So for people who missed our last update, that was called Zeus. So we're following these Greek gods as we progress here. And that was what we called Smart Search. But technically, you hear this term in the marketplace quite often, RAG, which is an acronym for Retrieval Augmented Generation. So we were talking previously about that Image Search and Concept Search, converting that information in your HawkSearch index into vectors. And then that's how you have these new experiences for search. That was that kind of first release. It was Zeus, and it focused mainly on RAG. But now with Athena, there's two particular things we want to focus on. The first one is what we call Seamless Search.
Because one of the things we want to make sure that we support is that all the capabilities in HawkSearch that are driven by what we call merchandising, things like campaigns and synonyms and other aspects that help drive that relevancy, as I was talking about before, we don't want to forget that. We want the campaign or whatever that merchandising rule may be to be continued into the smart search capability. So we call that seamless search. It means the facets that are returned, the campaigns that are returned, that they're still relevant even if you're doing a concept search. And we'll cover this in a demo shortly. But the other thing to highlight, though, is Gen AI. So AI, it's a very broad concept. RAG is one part of that AI, generating those embeddings, the vector databases we were talking about before.
But Smart Response is our Gen AI user experience, which means we can take a look at the content that comes back from the search results and then synthesize it, summarize it, compare it. We don't per se care what it is. The Gen AI aspect gives you that creativity to take the data that comes back from the search result and present it in a new user experience. We call it Smart Response, but it's leveraging Gen AI, but it's working off the data that's coming out of your index. And then finally, as Hermes, that's what we're working on now. That's what's coming out later. And we've been kind of calling it at the highest level a unified search, meaning right now, the three searches we've been discussing, the Image, the Concept, and the Keyword Search are indeed different search bars.
When it really comes down to it, we're going to unify that into a single experience so that all of the results appear in one response. You can, through configuration settings, determine how you want those responses to come back. I'll highlight one more thing here, and then Kyle, you can jump into some of the specifics. Just to spend one more moment on Smart Response and what the market is saying, and especially the smart analysts at companies like Gartner. For people who don't know, we did very well in a recent Magic Quadrant from Gartner that talked about search and product discovery and e-commerce. We were super excited about the results. What was so exciting about it is some of the capabilities that we're talking about in Zeus, and of course, now with Athena, weren't even part of the platform.
It hadn't been released yet before that Magic Quadrant came up. So the fact that we did so well even before we had some of these results is pretty significant. But the other interesting insight that's coming out of Gartner is very specific around Gen AI. And this quote here is saying, what you typically understand to be that search experience, the search bar, kind of that area within search, that it's going to be disrupted by Gen AI, which means it's going to be more of a conversational experience where the search is coming back with a question or a comment. Or, as we were saying a little bit before, this idea of synthesizing the response in order to engage in that conversation. And that's what we're building out.
When you hear this kind of conversational UI pattern, that's a fancy way of what we refer to as smart response. Kyle, any thoughts on your side? You see this stuff kind of with customers as they're trying out these features. Is this what you're seeing too from what customers are asking for?
Absolutely. In addition to what's happening in market trends and whatnot to help drive our roadmap with HawkSearch itself, a lot of that actually does come from our customers too. They come up with a lot of really good ideas. We take those ideas, build them into the product roadmap, in addition to anything that's helping drive from the marketing trend perspective.
Totally. Perfect. Speaking of it, Kyle, why don't you jump in? Tell us a little bit about Athena.
Absolutely. So we have the Athena release, which is actually already out. I know we talked about this last week, and it was just about to be released. But Athena is officially released for HawkSearch. It's our latest release. And I'll go through a couple of the details of it. And then, John, I believe you have a quick demo that we can actually show off.
Oh, for sure.
Yeah. So we'll get into the nitty-gritty here. So the first one is updated design. We're constantly evolving the look and feel of HawkSearch. And with Athena comes a brand new front end and brand new dashboard. So that's very exciting. We have Smart Response, which you touched on just a little bit earlier. But we can generate some information back to your customers after a query. This is totally customizable. There's a lot of use cases for it. We'll go over one later. But there's a lot of really cool ways that we can start to leverage this Smart Response. Seamless Search, the same look and feel across Smart Search or Keyword Search. So really trying to marry those two together. This one I'm really excited about. We're offering unique URLs per customer for API endpoints. So this is the vanity URLs for reporting.
So each customer will have the option to have their own dashboard indexing recommendations, search and tracking endpoints, which will then feed into our E360 dashboard for some really cool and advanced reporting. Alerting, so we have some advanced index reporting, some advanced triggering, lots of cool features coming out there. Then finally, we have one of our coolest new features. We are releasing the ability to create your own engine with products already populated inside of our development environment. That can all be done from the developer portal.
Totally. You know the funny thing, Kyle, I know you're in the interface all the time, obviously, IM demos or otherwise. And still sometimes I'm like, oh, wait, am I on the wrong screen? I forget how clean the interface is. So really pleased with it. And we have a great design team always making those enhancements.
Yeah.
You want to spend another minute on seamless search, Kyle?
Yeah, of course. And just to go back to the updated design, I think the coolest part about it is we're still keeping the kind of core functionality of HawkSearch the same while maintaining a brand new look and feel. So that's one of my favorite parts about it. There's a couple of things that catch me once in a while, but it really feels the same as it did before, just with a much better user interface. All right, and seamless search. So with the Zeus release, which came out a little while ago, two months ago now, I think, we pushed out some really big changes to HawkSearch to include image and concept search. With Athena, we're marrying the functionality of keyword search to these brand new features.
So regardless of how the user wants to search, they will have the same functionality across both keyword, concept, and image search. So all your rules that are set up for keyword search will also be honored in smart search. And you can see some of the merchandising specifics that will still stay the same, whether you're in keyword, concept, or image search, which is really cool. So that's what seamless search is all about. And look, your video automatically loaded here, John. So that's great news.
Oh, yeah. Yeah. So this is what we'll go through when we get into the Gen AI aspect. Think of it just as it's another search experience. Of course, you have the search bar, you have the search results, et cetera. Think of another experience that sits underneath the search bar itself that synthesizes, summarizes, or basically does anything with Gen AI based on the results that come back. We can be very flexible here. Of course, we do come with a default LLM, but you might want to have your own LLM, for example, for your large language model, for example. We can be very, very flexible around the prompt. When you get into kind of the details, how does this work? What's going on underneath the covers?
If you remember when we talked about the synonym generator in the Zeus release, that request went through a Bridgeline middleware that handles those details. It knows how to communicate the request. These are all now configurable on the front end. We're going to go through a demo that highlights this specifically. Why don't we wrap up with this, and then I'll do the demo, Kyle?
Yeah, absolutely. So in 2023, we took a huge step for HawkSearch documentation, of course, with all software, with really anything documentation is key. But last year, we did introduce some really cool features like the ability to send direct API calls to HawkSearch from our developer portal. We refined the documentation in general and clarified a lot of information around indexing. We included some visuals around it, all that good stuff. So we really took all of our documentation, put it in this beautiful little developer portal. And we also introduced the ability to make these API calls directly to HawkSearch from it, which was really cool. Additionally, last year, we released our fastest and easiest to use front end framework ever. And that was Rapid UI. With Athena, we're building off of these two HawkSearch staples. And we've now released the Quick Start Guide to automatically create an engine.
It also creates a front end framework. We have some pre-built smart search examples for RapidUI. There's new features also coming out to continue rounding out our developer portal. But in general, our developer portal is getting a lot of love. Quick Start coming out is fantastic. We can start having our customers or even prospects or whoever wants to come create a HawkSearch engine, see how it works. That tool is basically right inside of our developer portal.
Exactly. We'll go through it as well here.
Yeah.
Maybe I'll just quickly go through a bit of a laundry list. I do want to get to the demo. Two things I'll just highlight. You may have some thoughts, Kyle, too. The first one is for people who have heard about E360. It was an existing Bridgeline product. We've moved a lot of advanced reports there. But it was a little tricky to get from HawkSearch into those reports into E360. One highlight here for the Athena release is now the single sign-on. You just click on the button. It takes you in, takes you straight into your account. There's another one here, not related to E360, about more monitoring services. Or maybe you could think of it like threshold management. One of our enterprise customers kind of looks at the world this way, where the point is, great, you have reports. I understand.
We certainly get results back. For example, when we do indexing, was it successful or to track the results of that indexing. But it would be nice if you could monitor what's going on and send us alerts if there's a problem. So it's more like a threshold. If you see that the index normally has, let's say, 1 million records, and this one only had 800,000 records, let's say, tell me. Rather than me have to check and confirm that, just say like, hey, just letting you know this didn't match what you normally send in your index. The same thing with the stale index. Maybe, let's say, there is an issue. It's just not creating the index as you normally do. And it's been three days. It's been one day. It's been five days. The point is you could set some of these thresholds.
If they're surpassed, there's a problem. Like, hey, the index hasn't changed in three days. Send me that alert. Just kind of it's helpful kind of thinking ahead to, rather than me locking in all the time to see what the results are, you just basically tell me when something is out of the ordinary. Any other details here for you, Kyle, on the laundry list?
I mean, I was looking through our Confluence page with all the updates that are being released with Athena. It was hard to even start to pick which ones should make it onto this slide. A lot of little adjustments, a lot of cool things. I think we've mostly covered the most important ones, that's for sure. I am really excited about the E360 single sign-on, the vanity URLs, like I mentioned, where we can really start to track API usage and a multitude of other things per customer. Then, of course, the proactive index monitoring is fantastic too. I mean, we get excited about all the Smart Search and all that good stuff. I mean, reporting, developer portal, documentation, all that stuff is still super, super important.
Totally. Cool. All right, now the fun part. Wow, look at that.
Yeah, that lightning was incredible.
Yeah, that's how we feel about this. This is great.
I'm so excited.
Yeah. OK, so we're looking here at my account. This is why I was kind of commenting before for customers who may be online and seeing like, wow, it's a really new, clean interface, plus the nice greeting. Some of the things are on the smaller side. It's just a more modern way of presenting the dashboard itself and showing more of the information that's important to people. We always had this dashboard before. I admit I got very excited about adding WooRank analysis, which is another Bridgeline tool into the mix. I could understand that, like, look, just still keep it there, but maybe clean it up a little bit and provide a little more real estate for things like the developer docs and the user guide. We haven't talked too much, Kyle, about the user guide.
But it's also a great resource for videos and tutorials and insights into the platform, similar to the developer docs, but obviously for a non-technical user. Here's another example of kind of highlighting E360 again. Some of the updates are as simple as, you know, let's just clean up the top navigation where the information I need and kind of pointing me to these resources, for example, and even small details like this, you know, getting into that engine info, et cetera. Kyle, any other thoughts? You're in the dashboard all the time. You've already shared some. Anything else you want to highlight here?
No, I think everything that was reviewed is fantastic. And I think just to go back to the customer feedback component of it, because we did have WooRank kind of expanded out. And it was a little bit bigger before. But with some customer feedback, we adjusted it. And I think that's super important just to go back to that. But all the feedback from customers to help improve it over time, we really appreciate it. That's for sure.
Perfect. Cool. OK, well, let's get into some of the demos here. For people who have been with us before, this is our typical demo site. It is a catalog with outdoor equipment for hiking, camping, et cetera. We just have the site. We just call it the Adventure site. And for people who have kind of seen the search results, this is where the image search is, the concept search, et cetera. But what I want to focus on is that seamless search. So if I have a campaign that says, hey, we had a partner. And when people search for hiking shoes, I want to make sure I'm showing that campaign. So this is a very typical keyword search, obviously. But the point here is the merchandising rule understands that we have this campaign with Merrell. You can do whatever you want. It's all the creativity.
But where the real value comes in is, well, if I do the question, the lookup more like, I need some cool hiking shoes. The issue is, as we've covered in the Zeus release, some keyword search starts to break down. Notice the autocomplete isn't working anymore. We're running kind of into that issue that, well, we might not be returning exactly the right results, because this is more like natural language processing. This is that concept search. So if I then switch over to the concept search and I say, I need some cool hiking shoes, now I am returning perhaps a more interesting result set. But, and this is the key that I want to highlight here, it still recognizes that I have the campaign. So the merchandising rules stay the same way as they were before. And this is just one example of multiple merchandising rules.
Of course, you can cover other ones. But the point is the merchandising rule is still maintained even when you're using the concept search. The facets on the left-hand side are all maintained as well. So it just provides a significant amount of additional functionality there.
Yeah, and just to add a tiny bit to that, John, I think a lot of the reason that a lot of our historical customers have picked HawkSearch in the first place is the merchandising suite inside the Workbench.
Yeah.
Just blows pretty much everything out of the water. I mean, it's just so fantastic. And so being able to marry the smart search components with the merchandising suite inside of HawkSearch, it's really taking it to the next level.
Yeah. Cool. All right, now let's switch to Gen AI. So this smart response. So just bear with me one second. If you remember, when we launched the synonym generator, which was part of Zeus, if you remember, we had this interface that said, if you put in a word, let's stay with hiking shoes, because that's what we use a lot for our demos. If you remember this AI synonym generator, it was basically that first result or that first example of smart response. Of course, this is within the interface. These are our merchandising customers going in and making their changes and setting up these rules here. But this user experience here, just the way that it's laid out, and if you remember, underneath the covers is calling that Bridgeline middleware understands the prompts.
You could imagine the request that actually goes to the Gen AI system is passing in a prompt. This is a synonym, et cetera. And that's how this interface works. So the idea is, well, why don't we take this experience, this Gen AI experience, which we call Smart Response, and put that on the front end? So now we're looking at my demo site. It's the same content as the main outdoor demo site. But this is just stuff that I play with. So you'll notice I have the Smart Response here. So if I come in here and do the hiking shoes thing that we just saw before, number one, you get the results, as we were just saying.
But I also asked the Smart Response to say, look at the top three responses that come back and do a synthesis of it, kind of describe what's common across them or highlights about them. It doesn't per se matter. We're open to the creativity of what our customers are looking to do. But it's a new user experience that allows the visitor to have one more chance for a conversion. Maybe there's something that's interesting that's called out, something about the colors. There'd be links within here. It could look at the facets and then ask a question. This experience, this Smart Response experience, leverages all of the capabilities underneath the covers of the Gen AI that we're connecting to. But it's staying within a world that's in our index. This is coming back from what came back in the search results.
What I love about it is it's whether it's a Keyword Search or the Concept Search. So it'd be the same thing. I need the shoes, as we were saying before, would still come back with the same response. It doesn't matter what you're searching for, of course. It's looking at the analysis. This is one example of how you can leverage Smart Response. It's taking that kind of underlying technology that we had in HawkSearch itself before and is now making it available on the front end. This is where, as Gartner was saying, this is where that new innovation is coming in, what's going to disrupt what people typically understand just to be search, where I'm doing that analysis. I'm doing the synthesis. And/or I'm starting to ask questions, looking at what the responses are to help encourage that engagement.
Yep. And I think one really cool thing as well, John, obviously there's a million ways to really use this. But this can become almost like a new salesperson on the team. I mean, if you read through some of the response that actually was triggered here, I mean, it really helps guide the user before they even start to look at which products they might want to look at. It starts to help guide them before they even have to look at the results. And even if they pick, it doesn't need to be as long. Maybe it can be a shorter, sweeter message that is responded. But it's really a new little salesperson right below the search bar.
Absolutely. Yeah, great, great point. OK, I'm going to go back in here just for one second to get to the developer portal. Of course, you can get to the developer portal from HawkSearch.com. But this, as I said, is kind of a new opportunity to highlight things and where we're adding a lot of innovation here. So when I go into the developer portal here, first of all, for people who have been here, you'll see we've added a couple of new options here for people who haven't been here before. Kyle kind of highlighted this is where all our APIs are documented. You can test those APIs, et cetera. But I would like to focus on Quick Start. And this was something that was kind of always forefront for me when I started, which was, hey, can I just get an account?
I want to play around with it, et cetera. It took some time and needed to be configured and everything else. Whereas now, not only is there documentation that literally steps you through a process in a sandbox environment, this is actually just running in our dev environment, but with videos that help explain how the process works and what you can do. One of the challenges any software company has when they're trying to do tutorials and explain, whether it's a technical audience or a non-technical audience, is I understand conceptually what you're talking about. But how can I literally do it? I don't just mean, so I have access to your account so I can log in and do things. But I know what the data is.
So as a user, the guide that you're giving to me to step me through this process is exactly based on a sample set of data that we have. So if you step through this process, you get the account. What we put, and I don't think we highlighted it as clearly as we should have, this will be available to you when you say, just generate an engine for me. For now, you do have to ask someone, hey, can you please set this up, the engine? And then you can follow the rest of these steps. So what are those steps in setting up an account in HawkSearch after you have the engine? Well, the first one is, well, what are your fields? If you go back to the example, the outdoor demo example, it's a simple model.
It has the title of each one of your products, of course, an identifier for each product, the SKU, the description, things that you would expect are in the engine that we were just demoing before on that Adventure site. So you might say, yeah, but I want to create those fields. So you have two options. Of course, you could just log into HawkSearch and create them. There's an interface for doing that. But if you're a developer and you like APIs and you say, I would really love to run a Postman script. And by the way, we do give you a nice breakdown of all the Postman files here. So I should say all the Postman endpoints, one to create your fields, one to create your facets, one to create the index itself, et cetera.
We give you a nice Postman collection that steps you through each one of these steps. We give you sample data. Here's a CSV that you can run a collection against to create these fields. You can do the same thing with facets. Once again, if you want to do it in the interface, you could obviously log into HawkSearch and create a facet yourself. If you want to use the API, you could use the endpoint that we put into the Postman collection and run it against the CSV. Then you would need the data itself. Obviously, this data has roughly 2,600-3,000 products in it. We keep adding to it and cleaning it up. Anyway, you have a CSV with all those products in it. You run it against a script. You create your index itself. Now you have a page.
You can take that data and either use the original Rapid UI. For people who are once again a little familiar with this portal, this is the Rapid UI that we used to have when we just had keyword search. Of course, we've added for smart search. If you want it to look like this and have the image search and the concept search, anyway, you use this sample code. You could just even run it locally or throw it on an FTP server and share it with the team with the data that we give you here. Now, maybe the final thing I'll highlight, and Kyle, you might have some thoughts here, is as I was commenting before, we always had videos on how to do basic things. How do we do boost and bury? I heard your boost and bury is really cool.
How does it work? In this case, we have a video that shows you specifically for the data that comes in the sandbox with Quick Start, as opposed to conceptually telling you, here's how you boost certain results based on specific triggers or what it may be. This will work exactly the way we're describing it because you're using that account. Same with the campaign. That campaign that you saw before when I was kind of doing my demo, we literally show you how to do it. And then how do I update just the index I have now to include image search, to include concept search? You can do that straight from this video tutorial. So if you're interested in doing this, you're a partner, you're a little more technical, you'd like to understand how this works, just reach out to someone at HawkSearch. We'll create the account.
And then literally just step through this process. You can have your own site up in less than five minutes.
Absolutely. And I'll also just add super quick. I mean, getting the sandbox created is very fast. But even adding Smart Search and then even adding your client ID to the Rapid UI example that we have, I mean, it all happens pretty quick. I know our team has been, even some of the members of the customer success team, have been executing some of this work. And it all happens very quick. We can get through it very fast. And it's just very exciting.
Absolutely. Cool. Well, thanks, Kyle. I appreciate this. Thank you, everyone, for being here today. Obviously, click on this if you want to get some more info. Other than that, we're really looking forward to talking to you and in the future talking about the Hermes release.
Absolutely. I can't wait to see all the use cases that come out of the Athena release. I know yesterday we had a lot of cool use cases that we were talking about. Some of them got a little crazy. Excited to see what some of the customers that we have come up with and maybe even some other ideas. Certainly appreciate everyone joining today. Everyone have a fantastic rest of your day.
Great. Thanks, everybody. Bye.