FOBI’s AgenticBrain Launch: From Website Navigation to AI Conversation

Speaker A: Welcome to the Agoracom podcast, where investors discover great small cap companies. Thank you for taking Agoracom with you and make sure to follow our podcast.

Speaker B: Since the birth of the Internet, whenever we want to deal with a company online, we had to learn how that company want us to interact with it. We open its website, download its app, navigate through menus, search for the right page, fill out forms, and move through a series of predetermined steps. And that kind of navigation was great for a couple of decades, but users now want a much simpler and faster experience. Agentic Brain from Phobia AI is a solution that completely turns that model around. Why? Instead of forcing the customer to navigate the business, the customer can simply tell the business what they want through messaging channels that billions of us already use, like WhatsApp, iMessage, Telegram, and other platforms. Agentic Brain is designed to understand the request, connect to the company's existing systems, complete authorized transactions, whether that means finding something, booking it, changing it, authenticating it, resolving an issue, or completing a transaction. And perhaps most importantly for Fobi shareholders, Agentic Brain is designed to connect the technologies we've been discussing over the last several weeks. The Fobi Flywheel, fortress, fixer, alt ID 3.0, uh, and eventually Phobia's Next Generation Wallet. So here to talk about today's awesome press release is CEO Rob Anson and of course, CTO Yudesha Agarwal. Gents, welcome back.

Speaker C: Hey, good morning, George. Thanks for having us.

Speaker B: Labor Day yet, and you guys are banging out, uh, one press release after another, painting the Flywheel for us. So, Rob, first question to you. Uh, let's start with the problem Agentic Brand is trying to solve. We've spent the last couple of decades teaching consumers how to navigate sites like Agoracom, apps and everything I mentioned earlier. Why do you believe the next evolution is going to be much simpler, where instead of navigating a business, we simply tell the business what we want? Why won't people just keep doing what they've always done?

Speaker C: Well, it's about evolution, George. Um, we were introduced to, obviously, the, the Internet, and everyone built web pages where everybody taps and navigates. Then we introduced chatbots to try to accelerate some of the communication gaps, if you will. And in the case here now, this is the next evolution. As you've seen with Fortress and Fixer, the Flywheel in action. This is really, uh, what brings it all to life from, uh, an enterprise perspective. When we start to look at, uh, experiences, George, one of the things that everyone's been on about for years is personalization. And with this capability here now, when you have the integrated flywheel in action, this is all about personalization. It's all about action, it's all about efficiencies.

Speaker B: Are we going from, hey, where do I click to? George, give me what I need. A Gorcon website. Just give me what I need.

Speaker C: Yeah, you've got a direct path to action and execution. Now

Speaker B: Yudesha, we know you're the brain trust behind all this. All right, we know Rob's the visionary, but you're the brain trust. So I think one thing we have to make clear for investors is that, uh, agentic brain is not simply another customer service, uh, chatbot. We've seen that for the last several years, especially now with AI. So explain for everyone watching, listening at home, what's the difference between the chatbot that answers questions and an agentic AI system that can actually understand the request, access the company systems and complete the work?

Speaker D: Sure. Definitely, George. And you answered that question pretty much by yourself. A, uh, chatbot can only answer your question. It cannot get work done for you. Agentic brain gets work done for you. So here's the shift overall. Since 30 years we have been told, um, and we have been told to learn how interfaces work. If you have to change your flight, you have to go through 14 different screens of their app, which is totally new to you. This is where agentic brain comes into play. Can your business be talked to? Most companies would say no if you go to any company. If I ask to you, George, can your agoracom, uh, can be talked to? You might say no. Now how do we make your business talk to Agent Brain Agent. Once you integrate with agentic brain, your businesses can be easily talked to. And if you look at what's about to happen next, George, down the road in two years, most probably your AI agents would book the flight for you. And if a business could not be talked to, the AI won't be interacting with it. This is what agentic brain does. Agentic. Agentic brain makes the businesses being easy to be talked to. So now rather than going to an interface, clicking, figuring out how something works, you just talk. Hey, can you please reschedule my flight? Hey, do you have tickets available for the certain day? Hey, why is my network not working? The best part, it's um, it's industry agnostic, telecom, government flights, banks. You name an industry it can be integrated to. If you have customers and if you don't have a strategy, Agentic brain is your answer to this. AI strategy.

Speaker B: You make it sound first of all awesome because I, I look, I've even had clients tell me, hey George, I want to get into the Agoracom, you know, my, my gorcom discussion forum because I want to post a message my shareholders and they're saying, I wish there was a way for it to be much faster. But you know, right now you've got to go to the top of the website, search phobia, go to the phobia discussion for them, post a message. You know, that's just the way it is. I mean there is no faster way to do it. Um, you say in your, in, in your deck that I saw, uh, going from API to agent in hours. So explain what that actually means because it sounds awesome, but can you go like if I'm an airline, a bank, a retailer, other different industries with existing APIs, a lot of them have that. How does Folby turn those existing sites, uh, into agentic brain friendly? How long does it take? How easy or difficult is it to do?

Speaker D: Definitely, George. So you know, whether there's a very huge enterprise or where there's a very small business, all of them depends on very, uh, common functionalities. First is identification. Identification extends into authorization and authentication. Who this person is. How are we identified? We are identified mostly through our, if we, if we have logged in, we are identified through our emails or phone numbers. That's the one, that's how we are identified. Uh, once we are identified, then there are set of actions. The set of actions barely exceed hundred actions. The all the sophistication comes in, the security comes in, the speed comes in, the sandboxing, uh, of things. But if you drill them down, the actions are always less than 100. Consider an airline. What do you do from an airline? You check whether an airline is valid or not. You check what seats are available or not. You check their price, you check their timing. You can't even name more than 15 actions about it, right? So, so all those APIs which what we imagine as, oh my God, there's a complex infrastructure. All of those as a funnel comes down into maximum less than 100 actions. That's exactly where agentic brain sits. You plug and play agentic brain. Every company has documentation, every company has onboarding, every company describe how their APIs work. At the top level, you integrate agentic brain. It learns about your system and then it exposes itself as a chat to customer. So rather than George putting it his email and password and comparing him with a database, now gets authorized through his phone Number. Hey, this phone number is George. Who's George? George is a frequent flyer. Where does he fly from and to? He flies mostly from Canada to Greece. Okay, so uh, what does George would need? George needs a flight that uh, goes from Canada to Greece. When does he need it? So all of this is becomes easy by agentic brain. Because it sits within your system, it can call the API securely, it can retrieve the data securely. Ultimately the heavy lifting of clicking fetching is now replaced with this brain which talks with the customer and still fetches how generally the benefits do. So you don't have to change anything. There is no new forklifting. It's a simple integration with your existing APIs.

Speaker B: Anybody who would be looking at my, my cheat sheet here, my interview cheat sheet, my notes out in front of would think, oh, we pre programmed because my next question was going to be and you answered it. What about the user? How are they authenticated? So I love it. It's as simple as, because I thought maybe I, I've got to go to the website and I've got to sign up with a special thing and then they got to identify me. It's just off phone number. So as long as the airline has my phone number and they've got a gentic brain built in, is it, do I have to do any steps? Like if, let's say Air Canada and by not implying that Air Canada is a customer of yours, but if Air Canada is using agentic brain, uh, do I just simply at that point, uh, how do I use WhatsApp? Like how would I today ask Air Canada to find me a trip to Greece?

Speaker D: Yeah. So as a part of integration, Air Canada will be allotted a phone number on WhatsApp, on iMessage, on Telegram, also a username and you always text that number. So if you want to contact to AI Canada, you contact them on WhatsApp. Rather than a person sitting behind a computer replying you, it will be the agent brain that will be replying to you.

Speaker B: How long does this take to set up guys? So you score an airline, American Airlines. How long Rob, does something like this take, uh, to spring into action if an airline calls you tomorrow?

Speaker C: Well, as you just said, instead of months and weeks, we're talking days.

Speaker B: Have you guys had conversations? I'm assuming you know this isn't just coming out of the box, uh, today for the first time, I'm gonna assume you guys have had conversations. What kind of feedback are you hearing generally?

Speaker C: I think like anything, it's kind of a knowledge base to start, um, Understanding the art of what's possible. Uh, with it, um, as I said, everyone's looking to become more engaging, more interactive, providing a more streamlined user experience, understanding their customer. We talk about this all the time, unknown to total known customer. This is really the output of that, George. It's the communication that's embedded that people are using every day. Um, and this is exactly where everyone is looking as to costs plus and uh, becoming more efficient and the point of differentiation. I think we're going to see a lot of, uh, not just us, of course, but in the market itself. I think you're going to see a lot of, um, a lot of advancements. There's a lot of companies now which traditionally I would have anticipated being six months to a year cell cycle. It's kind of like let's roll the dice and let's see what we can do here. Because everything's moving so quickly. We, we don't want to be left behind. Like businesses can't afford to sit on their hands. Um, they become irrelevant literally overnight.

Speaker B: I would think, for example, the first airline. Let's assume again, I'm just, I'm using airlines example. If you know, George Carm. Airline is the first assigned with you. And users know that they can just use WhatsApp to book, confirm, make changes. Every other airline has no choice but to sign on almost immediately because they almost become irrelevant. If the user says, I can just do it through WhatsApp, why the hell would I go do. Is it. You think that's what we're talking about, Rob? Like that kind of fast transformation?

Speaker C: It is, um, you know, businesses always see monkey, monkey see monkey do you know, keeping up with the Joneses, as they say. Everybody's always striving for that next level of experience and you know, the customer acquisition strategy and narrative that's going to draw more people there. I think one of the key things also to George, to recognize for businesses and I'll use content. I don't really care how you engage with me, whether that's through web, chatbot, you know, call center, a genic brain. All I care is that you're consuming and doing business with me. This just takes it to a whole new level of obviously experience and efficiency.

Speaker B: I want to stick with that, that part. Every look, every industry's in play. I don't know any industry that this wouldn't be in play for. So the good news is you could walk into anybody on the planet. The bad news is you could walk into anybody on the planet. So is there going to Be verticals. You're going to focus on government healthcare events. Like what do you guys think in there? Or is, are the requests coming in that you're just taking all comers and stock them and rack them?

Speaker C: Yes, that's a good question. Um, a lot of this George, is really about um, I guess integrated opera. When we start to look at the various components of the flywheel, we can be all of it or we can be a part of it. Um, you like to use the word widgets and stuff like that. This is really the next level of widgets, if you will. Um, I'm very focused on channel partners. Udash and I have no interest in driving around the white van, the phobi van and selling this. Uh, for us we, you know, our scale will come through channel partner enablement. And um, the response here has been exceptional. Uh, I would say so far there's a lot of, um, a lot of, I would say jockeying and positioning within various industries as to how can we keep this to ourselves and then that doesn't really work. Right. Everything's moving so quickly that we're into servicing everyone. Um, for us this is built to scale. Everything you've seen that we've come to market with George, traditionally would have been half a year to three month integration cycles. Now it's literally hours and days.

Speaker B: Jesus.

Speaker C: So this is, yeah, I'm just going to say this is, this is the power of the roadmap of Udasha built here. You've got very sophisticated tech stack. As I said for many months leading up to this. We're probably one of the most relevant tech stacks out there quite honestly. Um, but more importantly George, the beauty about what I love here is, is this is an easy sell and that's where we get to here in the fall of building the channel go to market. It makes it uh, extremely scalable on our side.

Speaker B: Yudesha. It seems like the one component that must be necessary is a website has to have APIs and you'll explain APIs to everybody but essentially they're connectors that can connect to different places in order to perform a function. So uh, bridges, like websites have to be built in with bridges. So does that mean that this is more of an enterprise, like uh, an enterprise product? Are you going after big government, big industry, things like that? Like what happens to a website, uh, that doesn't have apa? Because a lot of small medium sized businesses just have websites. They don't have any APIs, they don't have the bridges and tunnels built in to connect. So does this mean this is enterprise, or does this mean if Georgecom Widgets doesn't have APIs, I can come to you guys say, hey guys, can you redo my website with APIs? Because I really want to use Agentic Brain. How does, how does, how does that factor? How does that play out?

Speaker D: Yeah. Yeah. So at the level of it, we have built it for enterprises for companies that do have APIs, because API is what acts as a food for Agentic Brain. That is what Agentic Brain feeds on, learns on, and connects with. And to be honest, we don't really care about the small website because one, there is, uh, not too much of revenue and then there is not too much of traffic or consumer there as well. So right now our focus is solely on enterprises.

Speaker B: Um, talk about how defensible this is. Either one of you can answer it, but you guys talk about Phobia's triple moat, private native and closed loop. So I guess, like the question I'm asking is what prevents, uh, another company from, from building this out? And obviously there's always gonna be competitors, so that's, that's okay. But you think you guys got a really strong moat. Can you, can you guys explain that as to what the difference is in your moat? And George is trying to build this out himself also.

Speaker D: Flywheel. It's the Flywheel, George. Agentic Brain. What it runs on. What powers the intelligence layer, Fortress. What powers the authentication identification layer, Alt id? What powers the consumer understanding support layer, Fixer. Well, uh, you want to build it, feel free, but then you have, you would have to invest years of intelligence, uh, uh, trillion of data points that Forbi has collected over all those years. I, uh, would say hours and hours of compute on which we have trained all of this. And then the vision that Rob brings and the technical abilities that I bring combining all of that. So that's why it's uh, like, uh, if any company wants to build it for themselves, it's gonna take them a super long time come up with this. And that's the beauty of Flywheel.

Speaker B: Yeah, Rob, Sometimes it's easier just to sell one product. Hey, I got this product. You walk into an enterprise and uh, you know, you get the sale or you don't. So some people might think like, well, now you got Fortress, Fixer, Alt id, Agentic Brain. Does, um, it make it easier to sell the Flywheel? Because now you're walking into someone saying, I just get it all for you. It's, it's almost like you're treating it's just one big product with four elements in it, right?

Speaker C: Yeah. I think the key there George, is really about future uh, proof. We're providing the gateway to transform, um, standardized business practice of today to now, tomorrow and futures, uh, evolution here. Um, I think that when we start to look at the, I would say the challenge of what many may perceive as oh, you're boiling the ocean, uh, these are all interconnected. So it's very easy to say would you like a sunroof, power windows, all these a la carte things. People would rather deal with one than four. So in this case here we provide the, obviously the, the new standard, if you will, of technology integration. But we provide all the comfort of future proof readiness.

Speaker B: Do you have and you may not because you may have thought about this or maybe do you get for everyone that watching at home? Because look, it's, this is very easy for us to understand. I think people are really understanding it. But examples always help. But do you have a simple example in which a widget company or some corporation would incorporate the entire flywheel and how that would look like uh, you know, just give an example of hey, you start with Fortress. It would do this fixer does that. Like, do you have an example like that that you could give to everybody to really picture it and understand it.

Speaker D: So since we are talking about airline storage, let's take the example of airline itself. So imagine if you can verify yourself uh, at the airline, uh, about the KYC or identification through Alt id. So that integrates the Alt ID part. Even when you want to change the uh, let's say the flight thing, right? Rather than opening the app and identifying yourself all again, entering your password, then validating by your phone, you don't have to do that, you just have to text the business. That's where the component of agentic brain comes into play. Now let's talk about Fortress, where Fortress comes into this. So airlines also need to have dynamic pricing. They already do have dynamic pricing. But how we can improve that? Whether this uh, based on the uh, speed at which the seats are being booked at the tourism, Whether we should price this uh, uh, plane at a certain price range or reduce that price, all of that intelligence, right now they outsource it. They have built a system that consumes their data. Whether you use OpenAI model, whether you use cloud model. With Fortress, it runs on their GPUs. It, sorry, it runs within their servers, it runs within their offices. So that's where the Fortress comes into play. Fixer is very important in these cases. Because customer support, like someone lost their bag, they have to wait on hours on the call. Fixer is something that comes into play. So fixer for customer support, alt ID for verifying if this person really is who they claim to be, whether they are at the airport or are they uh, spoofing, scamming somewhere else, uh, if someone wants to interact, book, change, book food within the flight. And finally uh, the fortress layer, the all the intelligence layer behind the flight. So all these four components comes into play with airline. However, based on the business, there could be more weightage on a certain component and less weightage on a certain component. For example, if you talk about a law firm, there would be a huge weightage on uh, Fortress, but less weightage on the fixer side of things. So based on the industry we interact with, we need to fine tune of what weightage we need to give or give on what product. But I hope that answers your question.

Speaker B: It does. And now we're about to go one level deeper into the matrix. So hopefully everybody, most people at home are understanding what you're doing, how this works. Let's go one level deeper right now. Up until now we've been talking about human messaging. So George needs to message the airline to book a ticket to Greece and how it's, how, how um, agentic brain is going to deal with it. But we're moving towards personal agents. Personal AI agents where you me rob people watching at home. Aren't you gonna have AI agents that said, hey, I want to go to Greece this summer, I want to stay there for two weeks, I want this and that and a little island, little mainland. Go book me a trip that's coming. Um, what happens when my AI agent starts talking to the company's agentic brain instead of me doing it myself? Uh, are uh, we talking about agent to agent talks through agentic brain?

Speaker D: Yes. That's one of the biggest benefits of agentic brain. If you go like companies today don't have an AI strategy. When you, it's so funny when you ask them about their AI strategy. They talk about using AI for content creation. They talk about using AI for discussing brainstorming solutions. They never talk about the thing which directly solves a problem and increases the revenue and makes it easy to be exposed to a wider range of audience. That's where agentic brain comes into play. And as I said in the opening, like today, you book your flight down the road, in two years, your agents will be booking the flight. Imagine two scenarios. A company with agentic brain and a company without agentic brain. So if you tell your AI agent to go book a flight and if it chooses an airline that does not have an agentic brain, in that case, your agent will open the web app, it will manually click on the elements, click the buttons. That whole process consumes a lot of token because your AI agent has to see what's on the screen. And reading an um image consumes a lot of token. Plus it's super slow. While it books your flight, your flight will already take off with agent brain. It does not have to click anywhere. Ah. One agent communicates with other agent. And very soon you will see AI agents are going to have their own language. They are already experimenting with it. So your AI agent talks to the agentic brain of an airline in a language they both understand. It could be English. Definitely works in English. I'm talking to you about the bigger picture here. They talk with each other. They sought a plan, and your agent tells that, hey, boom, this works. Your agent talks to a friend, talks to a guy that your agent knows, rather than learning a whole new interface and, and wasting a lot of memory time and token and money.

Speaker B: So Rob, you got about two minutes left because the good news about all this is you guys are hard booked on calls. So beforehand when we used to freestyle during the stealth mode, we don't have that luxury anymore because you guys are so busy. So really quickly you lay out three potential revenue models. Enterprise licensing, usage based scaling and closed loop performance fees. You want to talk a little bit about that and what the difference is?

Speaker C: Sorry, what do you mean, George? What's the difference between what those three

Speaker B: different types of revenue layers. Enterprise, uh, licensing or usage based scaling or closed loop performance fees.

Speaker C: I mean, ultimately, the end of the day, we like to get to licensing. Licensing we see will be the biggest path to scale, most definitely. Uh, a lot of the conversations right now around what is an annual license look like, that's really where we're trying to get to. I mean, the cost of the tokens in these things right now. Udesh is, I don't want to say too much, but you Dash has really architected something to be very disruptive in that market. Um, I don't think anyone will be able to compete on price. Uh, as we get going, uh, we'll have huge margin gaps built into it that we'll be able to aggressively adjust as needed, uh, based on the configuration architecture of which Udash has built our network and systems today. So I think that's if that answers question we're focused on Licensing. Licensing is the, uh, biggest path.

Speaker B: Um, yeah, that does, because you have multiple opportunities, but you want to focus on one. It's great to hear that licensing is going to be it because that's probably the fastest way to scale this all out. Last question, guys. Before you got to run, you desha, when do you think we might be seeing the demo of the Flywheel in action? And, uh, for investors who are really, really want to see how this all looks visually, not just talking to them about it.

Speaker D: Yeah, we are working, uh, really hard on that to create up a demo, uh, that explains how the entire Flywheel integrates. And we are looking at a timeline after the Labor Day, very near, uh, after the Labor Day. And, um, we are super excited to show this to our investors and the whole world that what we have been building and what we are capable of and how we solve the AI strategy for any company in this world.

Speaker B: Gentlemen, congrats. When I read this press release, I was like, okay, this is awesome. I started thinking about how even a Gorecom might be able to integrate it. I actually forwarded to my team, I said, once this interview is done, I want to hear from you guys how we could potentially integrate this even for Gorecom. So, uh, when you get me thinking that fast, I can only imagine what kind of response you're getting from discussions you're having. So, guys, congrats. I wish I'd keep you on longer, uh, but, uh, but thanks for everything. And this is a great. This is a great. Another rollout, Another great piece to Flywheel. Uh, and can't wait to have you guys back on, I'm assuming probably after Labor Day to hear more about the demo and maybe any other products. As for all I know there are more products coming into the Flywheel I don't know about. But, uh, thanks guys and thank you very much.

Speaker D: Thank you, George.

Speaker B: Thank you, George, everybody at home, thanks for joining us. Have a great day. Have a great Labor Day long weekend. Uh, and we'll see on the Agoracom full before them where you can ask your questions and interact with the guys, especially once Labor Day passes and everyone's back to normal schedule. But till then, have a great day. See you next time.

Speaker A: Hey, guys, this podcast is over, but don't forget to help your company by liking it or even leaving a comment. And then don't forget to help yourself by following us on Spotify, Google, Apple, or on your favorite podcast platform. So you will never miss another great Agoracom Small Cat podcast.