Veritone, Inc. (VERI) Earnings Call Transcript & Summary

All right. Good day, everybody. I'm excited to present to you today with some of my esteemed colleagues. My name is Ryan Steelberg, and I'm the CEO and President of Veritone. And I'd like to say we're also introducing what we're calling Veritone HR solutions, which is looking at really combining the technologies and efforts of both PandoLogic, which many of you guys know and use and trust, as well as a pending acquisition, which we're hopefully going to be able to announce the close in just a few short days here of Broadbean based out of London. So one of the market leaders, obviously, in job distribution, aggregation and dissemination. And today, we're going to talk about how we are leveraging and applying state-of-the-art AI into talent acquisition and recruitment advertising. And yes, we will be talking about Generative AI and its impact and there's exciting opportunities in this space as well. All right. To kick it off, let me introduce my speakers who are joining me today. I'm excited to have Sumit Gupta, the CTO of our conversational AI technologies at Veritone, formerly at PandoLogic. And also Corey Hill, who is one of our lead product managers and product leads over our Generative AI solutions and R&D efforts here at Veritone. Excited to have you guys both here today. Say hello. All right. We're going to drill into them later and ask some exciting questions and have them speak to some of the applications that they've helped build and bring to market and then also get their perspective on some of these newer AI technologies as they're rolling out here in the market. So let's look at an interesting quote. Steve Jobs is quoted often. And I think that we pulled this quote from one of his more recent commensurate speeches at a university before his passing. But it's very interesting. And let me read it to you real quick. "You can't connect the dots looking forward. You can only connect them looking backward. So you have to trust that the dots will somehow connect in your future." It's really interesting. And I think what he was speaking to with a little bit more context was intuition, karma, his gut feel. But I think I would like to take a little bit of liberty with this quote and think about, imagine he's talking about data sets, right? And again, historically, you're right, you haven't been able to necessarily make predictions on how you connect the dots or where trends are going in the future. But by looking in rears now and harnessing AI machine learning, you can start to do that. You can start to look at these, what's happened in the past, start ingesting and analyzing tons of different data points to truly make predictions, right, going forward. So I think it's a great segue. One of the greatest minds who candidly helped build so many of the experiences and technologies that we appreciate and enjoy today. But now with the age of AI and really with Generative AI, the art of the possible is being presented to us, and we're experiencing it real time. So let's talk a little bit about the evolution of tools and technologies in talent acquisition. Job-based advertising has been around for a very long time, with some of the most of the earliest reported job postings going back to the 1700s, I think as early as 1704 with job postings in Bostonian newspapers back in the day. And as distribution of news outlets and other sources of information were distributed across our country and around the world, that ultimately led up to Yellow Pages and then subsequently newspapers at scale. As we've moved into the mid-'90s and through the early 2000s, the worldwide web radically changed everything. Obviously, having a proxy of websites and ultimately, job aggregation sites that were interactive, was a complete game changer. And it wasn't just the ability for hiring companies to post jobs, but it was also for the job seekers. There's a place, a single source of truth, whether it's local or national in scope that people could, and individuals who are looking for jobs could go seek them out and sift through and search through all these different jobs. What's interesting is as we made that transition from newspapers to Yellow Pages to ultimately websites, something else was underneath the current of all this, and that's the data, and not just the data of the information about a job opening or a job listing but the data of the action of the job seeker searching and looking at these different job opportunities and job openings. Back in the day, you couldn't track exactly who was reading the newspaper. You absolutely could not be able to track if that job seeker was looking at that ad or even had successfully even picked up the newspaper. But today, obviously, in sitting on the shoulders of a lot of what the worldwide web fundamental architecture provided and a lot of the early advertising technology solutions in the digital and Internet age, we now have a data set, a data current underneath supporting the job seeker ecosystem and the job advertising ecosystem that now we're really starting to harvest and leverage. And that really gave us the birth and the marriage between these legacy rich data sets, now digital and scope and in vast supply with the onset of advanced AI machine learning tools. And so we fast forward a little bit, we start to see some of the earliest implementations of programmatic job-based advertising as early as 2007. Now remember, a lot of what we're talking about here in programmatic job-based advertising sits on the shoulders of a much larger industry, global, brand-based advertising and display-based advertising. So a lot of what we're seeing here and taking advantage of and leveraging for our industry focused on talent acquisition, you have to be able to look and compare and contrast of the trends that occurred in the past, sound familiar from Steve Jobs' quote, about how we're leveraging these insights, these technologies, these tools, these capabilities from ad tech as we transition into job-based advertising. And now obviously, Moore's Law, we're really seeing the effect of that is now this -- it seems to be that the AI machine learning utilization continues to advance so quickly. And we're not just seeing better targeting and been better job searching and indexing resumes. But now we have brand-new exciting technologies such as conversational AI and ChatGPT and new derivatives of Generative AI that are not only being used across many different industries, from finance to media and entertainment, but also for us here focused on the HR space and specifically, the talent acquisition space. So let's see, first and get a little bit more insight on the flagship product that we brought to market through PandoLogic several years ago, the really true first programmatic job advertising platform, pandoIQ. [Presentation]

Very exciting. And obviously, I'm thankful for many of you who are watching today, our customers of pandoIQ, and you've been realizing and experiencing the benefits of this amazing programmatic AI technology now for years. But again, what makes it work is the data. What makes it work is the interactivity, the connection points between job boards, between job seekers, between application tracking system, ATS systems, and that's what pandoIQ does. It brings it all together and makes it very easy and efficient for talent acquisition and recruitment teams to find exactly the right talent they're looking for where those job seekers are at. All right. Let's keep going here. And again, I touched on a little bit, but the rise of AI and TI -- TA is really grounded in a transition in many -- very similar to many different industries of being a data-driven decision-making business. That helps mitigate human bias but also provides a tremendous amount of efficiency gain and empowerment for these talent acquisition teams. And truly, AI allows a team of really any size to continue their operations with improved ease, efficiency and ultimately, performance. But like all new technologies, when you are transitioning from a purely manual human on-the-loop process to human in the loop or in some instances, human out-of-loop processes like full autonomous AI-based programmatic advertising, you have to also appreciate some of the downsides or some of the negative attributes that this technology can provide. And that's why you're seeing a lot of government agencies in municipalities across the United States and around the world, making sure that they're looking at these technologies to make sure that there's transparency that the -- that we are appropriately analyzing and removing bias from our AI models and whatnot. So you're seeing an increase in a rise in legislation which is good, right? We want structure, we want order and for this to be a thriving healthy tool set powered by AI, we need insight and governance from our legislation and our government bodies. In addition, we also need to have explainable AI. It's very simple. So if I -- the joke back in the ad tech days was, nobody ever got fired when they placed an ad on Yahoo!'s home page. A manual posting of an ad, right, for display-based or consumer-based advertising went through a very similar transition. But when you go to programmatic, can you explain where your ads are going? Can you explain why the copy was written a certain way programmatically? Are you targeting a diverse group? Are you excluding a certain socio or economic demographic group? These are all things that have to be considered when you're building and deploying sophisticated AI-based systems and you have to be able to reverse engineer as the providing technology company to explain how things are working. And this is not just about for protection or bias but it's also just explainability. So for that local talent acquisition or recruiting professional, they understand that historically, when they placed a job on a certain job board or an owned-and-operated job board of their own company, they knew what to expect. They knew what data elements that they were expecting back. They knew the criteria of success. When you go to fully programmatic, you have to be able to explain and compare and contrast that now programmatic automated function as compared to the legacy human element. So explainable AI, the extent to which the internal mechanics of a machine or deep learning system can be explained in human terms, what I just went through in detail. Second, understanding the how and why behind AI means you can better control your tech stack and proactively comply to both national and local regulations and governance. And then this is a little bit of a Veritonian attribute, but we are very passionate for AI for good. In almost all of our products across different industries, from media and entertainment to state and local policing and even in the legal system, we are very, very passionate and believe strongly that AI can be applied for good and still let us run businesses. From transparency and policing to improving the hiring practices and overcoming language barriers, Veritone is dedicated to enhancing the human effort, the human capacity by democratizing artificial intelligence and standing by pillars of compliance, transparency, explainability, key tenets for Veritone in all the products we build and deploy and sell, but especially in the talent acquisition and HR sector. Also, this is not just about our focus and attention on the hiring company but also on the job seeker. At the end of the day, this is an ecosystem, right? Ultimately, companies that are looking for talent need to find and engage and embrace the job seeker. And subsequently, those who are looking for a job, we need to help them find the right type of company and the right position at these respective companies that they want to work for. We've learned a lot over the last many -- several years and some of these are obvious, but I'm going to try to tie it back on how AI and machine learning takes advantage of these learnings. As you -- as we were discussing on the previous slide, transparency is critical, not just in transparency and how the AI models work, but also transparency in terms of the communication exchange, right, the contract between the job seeker and the hiring company. And I'll cover 2 very basic attributes where it truly translates into a significant improvement of performance for job-based advertising. One, historically, job-based ads that included pay information, compensation information as compared to those that did not, we've seen a radical improvement in conversion rate and a significantly lower cost per applicant when these hiring companies, when these job listings and job advertisements include compensation information. Seems pretty obvious. But again, how do you manage this at scale? Now we can analyze millions of job postings across thousands of job boards and websites and social media sites to truly, in near real time, find out what ads are working, what copy's working, what imagery associated with an ad is working and not just wait days, weeks or months, to get those learnings, we can harvest that information in neural time to improve our targeting models and make sure that we are putting the right ad with the right copy in front of the right job seeker at the right time where they are at. A second learning, and this is obviously very cute for many of us since the COVID days, is remote versus work from office. This is a radically changing dynamic that we're all dealing with right now, even us here at Veritone and our respective companies, but this also plays into job-based advertising. Similar to providing transparency on compensation, providing transparency and being clear and articulate about whether these are jobs that can be worked at remotely from home or virtually or of those that are work from the office is very important. When there's transparency and clarity, again, you're seeing, depending on what that person is looking for, significant improvement in terms of lowering cost per click and lowering CPAs and conversely, conversion rates. Misleading that user and when they start their journey, that job seeker, when they start that journey and they find out that a role really wasn't available for remote or vice versa, can have a devastating effect on the performance of your campaigns. Again, pandoIQ and a lot of the programmatic AI-based solutions, we help eliminate those mistakes, we help -- and optimize for, again, putting the right message in the right job in front of the right user to maximize the return on our job hiring investments. Now with the new lens of AI supported with all of this data and intelligence, we can reimagine what the hiring process looks like. Starting from attraction, job sourcing, finding the right candidate where they are, to engaging that candidate, right, whether it's in-person or through a chat bot or a conversational AI, which we're going to talk about, and then ultimately, through that journey of actually successfully hiring them and onboarding here. But again, it doesn't stop there. Ultimately, we want to look down the funnel over time. We want to be able to assess the quality of our applicants and be able to refine our job acquisition technologies and models, so even though we feel that we're hiring a great number of new candidates that seem to be qualified, we can answer questions like why are they leaving after 6 months? Or why is this the wrong fit culturally for our company? Too many of these areas are disparate functions. And at PandoLogic and underneath Veritone, we are very passionate to connecting these dots to not have just a linear sequence but an ecosystem, again, a self-learning opportunity that will take these learnings to continue to improve every aspect of not just the talent acquisition life cycle but the entire hiring and HR journey for these applicants and these employees. So we went through pandoIQ, the AI-based programmatic advertising platform. And now let's cross over to once we've connected with a potential job seeker, we want to engage that user in the most cost-effective and dynamic way. We want to introduce conversational AI for talent acquisition. Let's watch a short video of what Veritone and PandoLogic have brought to market and are continuing to advance upon. [Presentation]

Very, very exciting. Obviously, we're seeing how we're transitioning now from AI-based programmatic job advertising, really at the top of the funnel, to now engaging that applicant as they move through the journey in the hiring process. With that, I want to reintroduce Sumit, who heads up the Conversational AI platform and product suite at Veritone. Sumit, how are you today?

Great. Thanks for having me on, Ryan?

Yes. Sorry, I was a little long-winded in my intro to finally get to the real meat of the presentation, but I'm glad you guys stuck around for the presentation.

So we've -- chat bots and conversational AI in the area of recruiting. We've heard a lot about it. Can you explain in your own terms and words what makes for a better conversational AI experience? And how are customers evaluating performance or success with this tool?

It is a tough question to answer. If you just take conversational AI on its own, I'd say, you know this, Ryan, the team and myself, we've been working on conversational AI for about 7 years or so now. I'd say the first couple of years, I cannot tell you how many hours were spent sitting around just talking about what makes a good conversation and getting into the philosophy of it and then getting into how do you make this objective because it's a very subjective topic, one person's good conversation is not someone else's. So we spent a lot of time doing the traditional things, I'd say, around creating metrics around engagement, how are we engaging people, how many people start conversing with us, how many people complete the conversation, so that typical conversion metrics and engagement metrics. What we found, however, once we applied conversational AI in the recruiting space, on our first application was a screening product. So somebody applies, we send them out an invite and say, "Hey, senior employee, love to chat with you" or something to that effect, get them into our conversational experience. What we found that mattered, especially to our customers, but also by virtue of which to us, was our impact to downstream metrics. And the metrics that matter to the business. So for example, a typical metric, which would be time to fill for recruiters, it's like the gold metric. How does that get impacted when you introduce a technology like conversational AI in the process? So that really honed in on how we thought about the conversational AI and how we optimize our conversational AI. So it became frankly, less about conversation for conversation's sake, even though we still wanted to engage people, make it really interesting to people and keep our NPS scores up, but what became very obvious to us is, hey, we needed to really tailor this so that we collect good and unique information so that recruiters can make their decisions easier, faster, better, right? So -- and you talk about customers, so I'll just bring up Randstad as an example and quick tease, we have a case study coming out. So Randstad has been a long-standing customer of ours in the enterprise chat space. So they have chat integrated with their ATS, they use great people, in this case, as a ATS/CRM. Somebody applies. We engage with them. We do the chat, we move all the data back, including evaluation data into their ATS/CRM and they're able to leverage it. The key study very specifically is around a department that's hiring recruiters. So they're a recruitment team using AI recruiter to hire recruiters, try saying that 5x fast. But -- so this is where we are. So what they found and they have amazing results, what they found is time to fill increased by 20%. Our rate is faster now, but 20% with them. And you just look at the other numbers they're measuring, and it's quite clear, their team productivity went up by 135%. And it's not crazy because what they're saying is they're saving 20 to 30 minutes with a phone call every time our conversational AI goes and screen somebody, right? So the numbers are fantastic, and those are the metrics that actually say, okay, this is good conversational AI in the space.

So if I'm seeing -- listening to those stats were very impressive, and particularly on the efficiency side. This has obviously progressed a lot from just simple question and answers to a much more engaging dynamic conversation. Because at the end of the day, I mean, these are people, right? They have different unique experiences. They live in different areas of the world. They have different interests outside of just the job they're applying to. So obviously, in that demo that we just saw, I obviously saw not just a very -- a limited structure Q&A, but I did hear about Generative AI and potential GPT interaction. Can you explain to me a little bit more of how we've advanced it outside of just a simple Q&A process?

Sure. And yes, we've been leveraging a lot of similar and older technologies all the way back to when these large language models didn't exist, in various ways on getting it done. But I'd say like now the tool sets we have are so powerful, and they really are changing the landscape. But yes, for example, one of the things we always wanted to do was not just do very simple like, hey, we're going to ask you a question, you answer the question and it starts feeling like form filling. So we had to make sure that we were beyond that. One part of it was purely CX, like what's the conversational experience around it. We have focused and honed in on that. But then there's a big multimedia component to it. Can we introduce videos in there, which becomes now an employer branding opportunity, which I'm sure Corey will get into that when we talk about this more. And then how do we do things like -- and I'm glad you switched to the slide. So generate questions, it's not just questions, and we see this differentiation between us and a lot of our competitors, which is that those questions are specific to the job. So a lot of chat bots out there are really asking a standard set of questions. It doesn't matter what the job is really, I'm just going to ask you these things, right? So we are genuinely tailoring the conversation for the opportunity that you, the talent, are interested in. And what you see on this slide is a small portion of the pipeline. So we created a DSL LLM for a very specific task. And in this case, was around generating questions.

And sorry to interrupt you, when you say you're meaning a domain-specific large language model as compared to, I'll call it, a general untrained, if you will, large language model like GPT-3 or 4?

Correct. Thanks for that. Yes. So one of the specific tasks here was question generation, but also, generally speaking, you can imagine that you may need to take your large language model and fine-tune it to your domain. Our domain being HR, the language of recruiting is different, right? So the tasks we do in recruiting are very specific and you might get a greater yield out of fine-tuning a large language model instead of just using off-the-shelf, you'll get a better result and all that good stuff. The specific part of this workflow that we're showing is describing where we create the questions. I'll back up a little bit because what we do, and I hope people understand this, we really do a full-serve kind of a product solution, right? So it's pretty much hands off. A lot of our competitors that don't do the wrote One Chat. What they'll also do is ask the recruiters to punch in the questions. So we're really trying hard always as a product team to take the burden off the recruiter, not add it. Here's another tool you got to go manage and nurture and feed. So we will typically set up a way to go get your job descriptions automatically. It could be a job, it could be a API, whatever. And then we go and understand what the job description is. We have a taxonomy so we pull this thing apart. This same understanding goes to pandoIQ by the way, with its various algorithms. So now it knows how to distribute the job in the right way and everything that we talked about. The other side of the coin is, now we need to know how to converse about the job. So again, if you think about this philosophically, also industry-wise, recruiters are trained on a specific-domain. You're a tech recruiter for AI, right? Or you're a recruiter for marketing and you know the buzzwords there. You know how to talk about it. And here we are, we're basically what would have seemed in the past as a dreamscape scenario, taking any job automatically and being able to talk about it automatically, which is phenomenal. So what we've done in our DSL, LLM of course, is train it. So you specifically generate questions for a given job description, and those questions are what we then use during the chat, during the conversation.

Yes. This is -- I mean it's so exciting, and it really puts in context to what a lot of us have been able to try to -- have been experiencing directly with like ChatGPT, but now I think that our audience and our viewers are seeing how we're fine-tuning this. And so we're taking an advantage of these newer technologies, these broad-based large language models frankly, that can tap into more the zeitgeist or the general conversation with an applicant and marrying that with a domain-specific large language model that we've successfully onboarded and integrated through our technologies to really have in really a single conversation, both a very specific and organized and structured conversation about the job. But in that same conversation, get to know about that person about other interests that may not be in our domain-specific large language model specifically. So I can see that we could start to collect information to learn a lot more about this person, again, not just in the specific Q&A about this specific job, but about them, how are they going to fit into the culture of the company, which is so critical and important.

That's true. And some of the recruiters' feedback even early on would be around this topic of being able to see beyond the resume, right? So this conversation and the data coming back helps them see the person beyond the resume and maybe find a fit elsewhere. So you're absolutely right.

Sumit, this has been great. I really appreciate you walking our audience through this. So to be clear for our audience, we have off-the-shelf turnkey conversational AI solutions tailored for talent acquisition. And one I would say is more, once it's configured, a self-service solution, very tightly integrated with pandoIQ, which is pandoSELECT. And I know it's very exciting, and we've been working on it, you touched a little bit on it with some existing clients. But in addition to pandoSELECT and some of our self-service solutions that are integrated with pandoIQ, our conversational solutions are also available through enterprise-based solution, meaning without integration with pandoIQ. Can you explain just the difference between Veritone's products for pandoSELECT and how that compares to, I'll call it, enterprise conversational AI solution that we offer?

Sure. So pandoSELECT really is meant for the franchise shops, for example, is a good way to put it, where they don't have an ATS or they're not leaning into their ATS for decision-making. So we'll provide the dashboard where, as you saw in the clip, it went by maybe quickly, but very typically, you can see the jobs you have, you can see the applicants for those jobs and you can see their profile, what was the conversation, what were the highlights of this conversation and all that good stuff. A lot of enterprises, however, really and teams in those enterprises really, really work in and around their ATS. So they never want to leave that. That is the source of data the truth so they need to work in that realm. And that's where the workflows lie. So the enterprise product really is about bringing the chat functionality into your workflow, into your existing workflow, into those ATSs where the data comes back and insights come back and everything. So decision-making processes still live where they could be.

Yes, that's great. And a little plug to our pending acquisition of Broadbean who have successfully integrated with over 100 of the most popular largest ATS systems around the world. Do you think we'll be able to take advantage of Broadbean's ATS integrations to power our enterprise conversational AI solutions?

For sure, and extremely excited about that acquisition and the team. They're just a solid, solid team and looking forward to getting all those integrations, and, yes, absolutely.

So a summary here for, again, our viewers is conversational AI is a fantastic tool that is modular in scope. You don't have to use it when you engage with Veritone and PandoLogic technologies. But it really can help you engage again that candidate where they are, when they're available, and as we're in a tease here in a little bit, potentially in the language that they want to hear it in. Thank you, Sumit, for walking us through this incredible technology and giving us some insight of where this is going. And again, at the end of this presentation, we'll have a URL up for those who are interested in getting more information about any of our conversational AI applications or technologies. And we teased a little bit. So now let's go -- let's look a little bit into the future and look what some of the newer AI-based models and technologies that are currently available and coming on the horizon here and how we can incorporate some of these into talent acquisition. So we talked about programmatic job advertising. And then we talked about conversational AI and how -- and how conversational AI for talent acquisition can be seamlessly integrated with the top of the funnel, job advertising but also, as we mentioned, on the enterprise side, fully integrated with the ATS systems. So it's very, very exciting. Now let's talk about Generative AI. And what I mean by Generative AI is the GPTs of the world. This is actually leveraging these tools to create new pieces of content, to create potentially a new experience when you're engaging a job seeker. And with that, we're going to focus, on for this presentation, on one aspect of synthetic content creation, which is synthetic voice, in-house synthetic voice technologies which can be applied for the HR space and specifically for talent acquisition. So let's watch a short video of what Veritone does and provides as it relates to synthetic voice in the industry. [Presentation]

All right. Exciting stuff. With that, I want to bring in Corey Hill, who is going to talk to us about what we're building here at Veritone, who we're servicing with these incredible new synthetic voice technologies and specifically, how these can be applied to the talent acquisition marketplace and opportunity. Corey, welcome to the show again.

Thanks, Ryan. Appreciate it. Glad to be here.

Corey, very exciting video. A lot of us have different experiences and exposure to some of the synthetic voice technologies out there. Can you walk -- just provide an overview of what you've seen in terms of some of the really practical and utility-based use cases of synthetic voice? And then maybe you can go on a little bit about walking through how this technology can be applied specifically to the talent acquisition opportunity?

Absolutely. Yes. I think we're seeing a lot of opportunity to reduce the amount of time specific talent needs to spend inside a studio recording booth. They're able to kind of replicate and scale their voice very quickly and still maintain that kind of brand experience that their specific voice and all the attributes around it provides. And so we're doing things all the way across the gamut from taking real-time sports data and producing commentary to generating very unique and specific experiences for personalized greetings through some partnerships we're working on with some companies in that space as well. And so I think the opportunities are extending the gamut across multiple media and entertainment spaces. But I think for the opportunities in talent acquisition, if you think about it, the acquisition component, we look back at kind of the life cycle wheel of talent acquisition that you showed a little bit earlier, and there's significant opportunities in each of those little slices of that pie, right? And so as you're thinking about the ability to produce recaps of maybe your presentations on a specific conference and give the job seeker the ability to ask questions about the current conversations that you're having with not only customers but specific social media posts as well, it enables you to effectively reproduce that content in a summarized way, but maintain that brand consistency. So if you have one voice that's presenting, maybe that's a recruiter, maybe it's your SVP of sales or marketing, you can take that person's voice, clone it as we saw in Anne's case, and reproduce that content in multiple languages so that you maintain that consistency. And that's really important, right, in the talent acquisition space is to reduce any bias and maintain that consistent experience across each candidate regardless of their localization or specific economic attributes around their location, too.

Yes. And I can see, again, a lot of opportunity to really automate and personalize a lot of the different areas of the life cycle for talent acquisition. I think in the industry, I think many of us know Chad & Cheese Podcast, and what do we do for Chad & Cheese, who a lot of us know in the HR and talent acquisition space? How are they leveraging synthetic voice, for example, their podcasting when they're talking about HR-based technologies?

Absolutely. I think those guys identified a significant opportunity to scale their presentations into multiple languages. So really taking that into Portuguese-speaking markets, Spanish-speaking markets and kind of providing that expertise and that just that really engaging conversation that they have around HR and present that in a highly hyper-localized format, right? And so instead of someone sitting there and potentially reading subtitles, they're actually able to engage in a natural -- much more natural way than if they had to kind of do some translation and listen in English.

Thank you, Corey. Now I can see how we're really helping close the loop on the end-to-end opportunity for talent acquisition, starting again with programmatic AI-based job-based advertising to managing the experience with conversational AI. And now with synthetic voice, we can personalize it and turn that engagement experience all the way from an audio-based job ad in multiple languages, but potentially now integrating any synthetic voices with and into a dynamic kind of next-generation experience of communication and interview process. So if I pull all these technologies together, I think we can now see when we combine Generative AI next generation with synthetic voice and now showing you here with dynamic and programmatic avatars, we really can change the entire experience from the top of the funnel to a pure programmatic opportunity for recruitment with not just generic chat, but with actual synthetic and personalized voices and avatars. Let's take a look. [Presentation]

I love that presentation. It's really interesting. So now I'll kind of open it up to both you, Sumit and Corey. What am I looking at here? What different technologies and different tech stacks do we bring together to show you this experience of -- and for those who may not be aware, the voice of the avatar is fully synthetic. So I'll turn it over to you, gentlemen. What am I looking at here? What technologies were brought together to create this experience? And how do you see this to be impactful and beneficial for talent acquisition?

So it is pretty impressive from a tech stack standpoint, right? So we've got a voice clone. We've got an avatar. There's a lips sync happening with the voice in real time, mouth movements, right? There's emotions that are reflecting that as well. The talent that's talking back to the questions and answering those questions are going all the way to the conversational AI. It's processing the responses. Knows what to ask next. So there's conversational AI on the back end. There's voice, there's text tech as well, and then there's visualization through an avatar all coming together. It's pretty impressive.

Very much so. I think typically, I focus on the voice clone aspect and the ability to take that individual's voice and now kind of in the pipeline, right? It's a full experience. So you can increase velocity, increase engagement by not just sending e-mails back and forth, right? You can have this person's voice now present summaries and kind of give you a recap of how the interview went, maybe give summaries of the people that are going to be interviewing with in the future, maybe they send up kind of a cool pump-up reminder like, "Hey, we've got this great interview coming up next week. We are so excited to have you here". And so think about the recruiters now being able to scale their ability to now talk to more job seekers, right? So it's really engaging you to not only just expand and find more talent, but really create an experience that makes people look forward to continuing that conversation with your organization.

The branding experience here is just so powerful. Imagine coming away from, I don't know, interviewing at Disney and talking to your friends, "Hey, Yoda interviewed me", although it'd sound weird, but -- but it will be a great experience, and it's a great branding effort.

I agree. I think employer branding and I'll say brand continuity is critically important. Again, now you can imagine an experience that you're investing in a synthetic voice or an ambassador of your company. It could be a robot. It could be your CEO, as we mentioned, or you, in your case, it could be Yoda, but maintain that all the way through the experience, right, from the job advertising to the conversational AI, to an interactive experience with a potential avatar, you can control the narrative. You can control the branding. You can control the messaging. Thank you, guys, for walking us through these exciting technologies and how Veritone we've applied these unique and dynamic technologies into very turnkey applications that our customers can start leveraging immediately. And again, Jeanne Meister, a very well-known author and kind of futurist in many areas in the HR space. And just -- we'll kind of leave with this quote. "AI will augment HR and give HR time to work on more strategic business issues. The opportunity is to use AI to streamline HR manual processes and provide a more consumer-grade service to employees." And I think if you look at the technologies we walked you through today, and more importantly, the off-the-back-of-the-truck and turnkey-ready applications that we walked you through today, I think businesses of any size can see how AI can truly impact and improve their talent acquisition efficiencies and cost. Thank you for your time today. And for those who want to get more information about anything that you've seen today, if you want to watch this video again in slow motion and try to take notes of some of the information we presented here today, please visit at veritone.com/aiforta again, that's veritone.com/aiforta. Hope everybody has a great day. Thank you.