The Trust Layer for AI Agents — Ravi Bijlani, KYXStart.Ai

SPEAKER_00: Most organizations are not behind on AI. They are behind on the thinking required to use it. I am Arnand Sipatnaik, ex Microsoft Data and AI executive, CEO of Tipsora, and your host. This is what comes next, the podcast where we talk about building the kind of organization that actually wins in an intelligence-driven economy. Let's get into it.

SPEAKER_00: The guest for today is Ravi Vijlani. He is founder and CEO of KYXstart.ai, where he is building the trust and verification layer for the agentic era. The identity infrastructure that lets AI agents act, transact, and make decisions on our behalf without breaking down the trust humans rely on today. He works at the intersection of AI, identity, and security and spends most of his time thinking about the gap nobody has solved yet. How do you verify who or what is actually on the other end of the autonomous decision?

SPEAKER_00: So with that, Ravi, welcome to the show.

SPEAKER_01: Hi Ramasu, thanks for having me here.

SPEAKER_00: Yeah. So uh we uh talk briefly. Um I uh told you uh I wrote this book on uh my framework uh called the durable uh advantage cycle. Uh so it's something that I have observed um over the years working in big tech and observing how other uh startups have grown to big companies. So it has uh five stages. First, you identify your niche, uh what you are good at that others would find hard to replicate, then you diversify, and then you build a platform, and uh then you build your team structure around that ownership mindset, and then you take it to the next level by investing into research and development and managing external environment.

SPEAKER_00: Right? So I kind of structure the podcast into those same stages. So I take you through those uh journey of those stages that I talk about in my book. Okay, so you started this company um you know with the idea of uh creating the identity layer for um AI agents to build trust to allow people to uh do commercial transactions uh using AI, right? Uh so you clearly found your mess. So walk me through uh you know a little bit of your background and what happened and where you saw this gap that made you realize that you know this is an opportunity, this app needs to be built.

SPEAKER_00: How did uh KYX start for you?

SPEAKER_01: Yeah, yeah. So uh I appreciate the uh the question, uh Nasu. So, you know, I've been in this identity and payment space uh or at that point when I started Kickstart for about a good 24 years. Uh we started Kickstart uh two years back. Uh one of the main things that we saw was uh, you know, AI. Again, there was an AI boom, everybody was talking about AI, they were really focused on AGI or path to AGI. Uh but somewhere in the middle, we thought, you know, AI is gonna be autonomous, right? In early 2004, there was no word like agent.

SPEAKER_01: Agent really didn't come into incarnation until OpenClaw came in and you know, the uh the word was, I mean, the word was there, but the word was really linked to AI during that time. But what we thought was, or what we believed, was autonomous AI is gonna come in and AI is gonna wanna make all the decisions. And like the the the the growth that we saw, what what I call it as a uh as as a poll growth, right? It's not a hockey stick, it's like a straight line that we saw, uh, what happened with Chai GPT and others, what we realized was consumers are gonna be a big part of it.

SPEAKER_01: And it's it's gonna be more B2C than just B2B, because that's what you know when when you think about a uh a prior version of what we thought was predictive analytics, machine learning, and in some cases, people were using some of the AI models. It was more about B2B. Oh, I'm gonna do this financial modeling, I'm gonna figure out this you know, credit model. It was never really uh uh B2C or or C2B rather, right? Consumer to business. That's the gap that we saw. And one of the gaps that we saw was that you know there is no trust layer.

SPEAKER_01: Today, uh Nasu, like a quick question, and probably you know your audience might find this uh uh uh you know rhetorical, which is when you open a bank account, you have to go through you know a ginormous number of steps, right? You have to figure out your phone number, you have to figure out your name, your address, you know, your driver license, and you know, many of the other things. What we thought, and and now I'm gonna use today's dunk, what we know is that agents cannot upload a driver's license, right?

SPEAKER_01: Because they don't have one. Agents cannot uh uh cannot really take a selfie because there is none, right? And so what we thought was that trust layer is needed where all these, you know, uh so-called LLMs or or any of those uh you know artificial intelligence interfaces would need a trust layer where they can say either this business is valid or this consumer is valid, right? And more importantly, this payments that is going through has been verified and it's been used with the right purpose in mind.

SPEAKER_01: So that was like the like the light bulb uh that we got. And the most important thing that we learned, and I think I believe we've done an amazing, amazing job of that, is that it's not geographically bound, it is gonna be global in nature. So today we build Kickstarter as as an as as a platform that can support identity verifications in 180 plus countries. We can verify more than 5 billion users, we can verify more than 400 million businesses, but more importantly, we can now connect autonomous agents to those businesses and to consumers to actually go and perform agent tech commerce or agent tech transactions or agentic money to uh uh money movement.

SPEAKER_01: But let me just give you this something to think about, right? 8 billion is probably give or take, 8 billion is the world's population. More than 400 or close to more than 400 million businesses, they're gonna spawn about, you know, let's just say good 400 or 500 trillion agents. Out of this, which which me, Gartner, and many others, Ravlin and many others believe, 25% of those are gonna actually do financial transactions. That's my time. That's why we built this platform because we wanted to be the trust layer for that level of uh uh financial transactions.

SPEAKER_00: Yeah, yeah, that's great. So, you know, explain me a little bit more. Like, you know, I I know like when the agents as a concept got created, you know, um I was at Microsoft and Microsoft came up with this Agent 365, like a dashboard for agent, and uh, you know, within the Azure Cloud, that there was um this concept of managed identity and service principle. And you know, so when I was uh advising security uh on AI and agents to uh customers, I I would always say that you know, when a user creates an agent, their uh identity or they have the option of creating a managed identity or a service principle and attaching that uh to the agent, which is like no different from you creating an app and letting um you know that app have access to things.

SPEAKER_00: So agent is just another app. So the same security principle applies where you know you either attach it to a service principle or a managed identity, uh and that's the identity, and uh that's how the you know, beyond that identity, all the traditional role-based access control uh kind of applies and the trust is established based on that identity, right? Now, explain to me that you know uh besides that identity, like the financial transaction, like how the credit card transactions happen, uh, you know, the fraud detection, all that also already exist on the finance side, right?

SPEAKER_00: So explain to me like how your identity and uh commercial trust, like as you're calling, uh how is that different from these service principal and managed identity-driven identity that we are used to for you know autonomous apps or agents?

SPEAKER_01: 100%. Uh Arnasa, so one of the main things that you will like connect the dots on is that agents kind of emulate your identity, right? Think of them as a as a shadow, but a performing shadow, so to speak, right? So you are right to say that that identity can be assumed by by whatever, you know, if they have an existing thing. The the biggest missing part, Arunasu, there is that that identity was a static identity, right? There was like, oh, I'm just one time and I'm done, right? And then it was led with friction.

SPEAKER_01: For agents, them to be, you know, A, uh they have to be, you know, uh they have to stay in a memory because they have to do multiple tasks. It's not like one task and you're done, right? In in case of the world before that, you were there, you were interacting, and you were doing, in this case, agent have to do that. So we we believe, and that we've seen that in real time, is that the world has transformed from a static identity to a signal-based identity, right? So you must have heard about customer 360, you know, all that stuff about a particular customer.

SPEAKER_01: Now all of that has to be emulated by the agent, right? So yeah, you know, we we were very used to saying bots are bad, 20 years, you know, don't do that. It's the same thing now, the bots are good. But there is a new term that came out, HNP, human not present. Car not present, now it's human not present, right? Uh, and and so for HNP transaction, in certain cases, you need human in the loop. So at that point, you know, bring human in the loop. So what we have bear is kind of a way where you can bind the agent to a device and to an identity because they both have to be connected at the same time.

SPEAKER_01: Earlier, the device part was kind of just make sure the device is good, it's kosher, because you still had human interacting. Now it's, you know, so to speak, device to device, system to system. So you also have to associate the device to human, that agent to human. And then what we have also done is put put a so-called smart contract around human and that that agent that, hey, if I tell you to go buy a flight ticket, don't buy me a$10,000 ticket because I'm gonna say don't buy a ticket more than$500, right?

SPEAKER_01: That's so that that's the base principle of it. But as the agent traverses and it does whatever it needs to do, right? I I get it within the enterprise, Arunasu, what you just said, within the enterprise, it has the guardrails, you know, there are simple systems that it's gonna connect, right? But think about an example where you're like, go buy me a ticket from Atlanta to New York, uh, you know, and keep it less than 500 bucks. At that point, a lot of things have to happen. The agent has to first get verified before it starts transacting.

SPEAKER_01: It has to be verified by the CDN, so the CDNs let that agent through. And then the receiving party, which is, you know, who whichever the airline of your choice is, or you know, it has to accept that agent. So we have built what we called as, you know, we call it KYCA, the world calls it as KYA. We call it know your customers agent, right? And then we are associating multiple uh information. We you can call it token, you can call it whatever that is, right? It's also an open source. So it's not like something that, you know, I came up with.

SPEAKER_01: The more important thing is that the token is verified, it is watermarked with trust, with the trusted identities. And then uh guess what are an assistance? Not only it has to interact with the with the you can call it uh commercial or commerce provider, like a flight ticket or buying me a you know a t-shirt or a headset or whatever that may be, but you also now have to interact with payment gateways. So you have to form that element of it, and those payment gateways also have to interact with it.

SPEAKER_01: And they have to interact with an agent, they have to be able to verify the agent, take the information out of that agent's token to say this is the credit card associated with it, right? So all that has to happen in an ecosystem. So and that's what uh really what we have built. And then the the most important part, Arunasu, is that you know this ecosystem has to exist with multiple parties involved, and we are part of that ecosystem that we've built together uh with the right uh uh you know participants, so to speak.

SPEAKER_01: And then what one is is the most important thing, right? You also need that agent registry because you need that auditability, right? Today, you know, uh I don't know how many people they drive autonomous cars. I don't want to again take company names, but if you have a Tesla, if it speeds, you get the ticket, not your car Tesla, right? So if you if your agent does a mistake, somebody has to help be held accountable and think about that linkage is one of the key components. And that's the trust layer that we've been talking about, or that we have built, and that's the trust layer, I believe, is is the most important thing for uh for the next phase of digital commerce.

SPEAKER_00: Yeah, yeah. So, so you know, uh it's it's a very novel idea. Uh can you double click on um you know your core of you know the idea? Because you're talking about a global standard, right? Like you by yourself, like you can come up with a way to do it, uh, but you cannot set the standard, right? So is there a global standard being set, just like we had for like you know, open ID and you know um SS authentication? Like there are authentication standards, right? So is there a governing body that is that is global, that is accepted?

SPEAKER_00: So is is there now a global standard for agent trust and um uh authentication? And you know, um, are you a contributor? Uh you know, who else is there as a contributor and how good is it that it it it can now you know officially become a global standard?

SPEAKER_01: And you know, uh that's the billion-dollar question, right, and also, because it is actually I take that back. That's a multi-trillion dollar question. So there is an institution called Linux Foundation, they have an AI foundation as part of Linux Foundation. A lot of companies are involved with that, right? And then everybody is giving their best. But I think I think what has happened, and this is something that's happening in front of our eyes, is that Google has created their own protocol. Stripe has created their own protocol, so has MasterCard, so has Visa.

SPEAKER_01: And and we don't know who's gonna win. We we nobody knows who's gonna win, right? It depends on whatever influence they have, they can enforce. Oh, if you want to talk to Visa's agent, you have to use this tokenization. Okay. So what we have done is that we put a layer of obfuscation on top of that so that you know these commercial entities can talk to those, their agents can talk to those sub-agents, so to speak, with whatever protocols that's out there, right? So we we don't know which protocol is gonna win.

SPEAKER_01: Uh and and that's the that's the key here. Uh, we are kind of uh maybe a year, year and a half into this, right? So that's to be determined, uh Al Nasu. There is no silver bullet to that. But what what you brought up is a very important point, right? Like the Kickstar cannot set the standard. The way the standards are being set would be, you know, something that is uh that is gonna one of the main things is gonna be which one takes off commercially, right? Which has the broader adoption that will eventually become the standard, so to speak.

SPEAKER_01: So yeah.

SPEAKER_00: Yeah, yeah. So so so so you know, you you brought up my uh next uh big point, which is like, you know, we're talking about big corporations like the MasterCard, you know, Google and Stripes of the world. These are multi-billion dollar companies, right? Yeah. And they have their own standard. And like one of these or a combination of these could eventually become the standard, right? Yeah. And you know, like if it does and when it does, uh, where does that leave you? So where is your value prop? Like, you know, are you competing with Google on standard?

SPEAKER_00: Or if Google wins the standard, where does that leave you? Where is the your uh market, right? What what is the you know uh market segment that you own through your service if you know these companies are in owning the standard?

SPEAKER_01: Yeah, so standards uh is gonna be around how the how the agents talk to each other. We are the trust layer on top of the agent. So, you know, any kind of standards can succeed. It really is gonna help us because then I don't have to maintain, you know, whatever, 10 or how many ever odd standards they are there, right? So for us it becomes an easier thing. Uh where we play in our NASA is that more about when you create an agent, you bring human in the loop, you verify the human, and then you bind the human to that agent.

SPEAKER_01: So how our agent communicates is is is really uh it really depends, you know, or or from our point of view, it's it's it's something that's irrelevant to the broader business that we are in. We can adapt to any any communication channel. The part where where I'm so excited about participating in the economy is that, you know, bringing trust, so or not, so folks like you and I can go to any any LLM that's out there and say, buy me the ticket, and you are okay with giving it your financial and of course, you know, your financial product, like, okay, you know, go and do this, uh, do this transaction.

SPEAKER_01: Let me give you like a parody, right? Why that is important. So, you know, this this is something I I I heard about and I don't remember where, but it's not it's not mine, so I'm not gonna take credit for it. So you think about Amazon, right? Why was Amazon so amazingly successful, right? Because they, when they came into play, you you know, it was digital web, right? If you remember, like at that point, everybody wanted to build a website and put a catalog. They didn't want to transact, but that part was there.

SPEAKER_01: The part was also existing where you could take the credit cards in a secure way and you could process some transactions online, right? Not all of them. People are not used to it. And then the shipping part. What Amazon did was it used the trust layer for all these three and created an ecosystem. So, of course, there is some Amazon that's gonna come out there that's gonna connect all the trust layers, and we are one of the trust layers in that case, because Amazon also uses identity verification before they let you create an Amazon account.

SPEAKER_01: So think about from that point of view, you will you will need that trust layer to to participate in there and then for customers to feel comfortable that, hey, I've been verified, my data is not gonna get you know leaked, so to speak, right? And and uh so yeah, so I I I I think I don't know how well that landed, but that's the analogy that's yeah, yeah.

SPEAKER_00: So yeah, I mean, you know, uh I I want to clarify that a little bit more, right? So so so today, like you know, if you look at the agent economy, right? The protocols exist. Now it's like you know, 50-50, either A to P or you know FCP, right? So those protocols exist, and the protocols are open source, right? Protocol anyone can uh implement, nobody owns it, right? But but in in in the economy, the like that is the data provider, the people who create the agent and people who own the LLM, so they own the commercial side of it, that we pay token uh cost to them to use it, right?

SPEAKER_00: Now, when when we look at this agent e commerce, right? Yeah, so whoever, who whichever protocol eventually wins, it's not the protocol that makes the money because protocol has to be an open standard, right? Yeah, it's like you know, saying uh how the internet works, nobody owns how the internet works. That's an open standard. Uh, but you know, if you uh Subscribe to Netflix, you have to pay subscription to Netflix, right? But um let's say there is an open standard for agent commerce, and that's open.

SPEAKER_00: Nobody pays anybody for using that standard. What is it that people will pay you? Is it storing their identity with you? Or like you know, are you kind of becoming like the Facebook ID or the Google ID of the world? Uh, that you know, if they have an ID with you, then they are trusted. And, you know, is there a subscription model to have that trust? So, what would be your monetization strategy in this case?

SPEAKER_01: Okay. Uh, yeah. So let me answer it in this way, right? One thing is that LLMs are very focused on building a better model. They don't want to play in the regulated space, right? So they're not gonna, I mean, at least from our opinion, right? Is that they the regulated space, the regulated space may not even let them come in because it's, you know, again, you know, one thing and then it it opens up uh a different kind of ones, so to speak, right? So from our side, how we do business that we work with other businesses.

SPEAKER_01: So, you know, the the time where you feel like creating and Kickstarter ID, I don't think anybody's gonna need any ID other than the ones that you use to log on to any LLM or any other platform, so to speak, right? What we are trying to do is that even get away from that. So today, let me ask you this question: Have you done any banking or any high-risk transaction without your phone being involved in it? Like either you get an OTP or your scanning something, you are there, right? So now let me ask you the next question.

SPEAKER_01: Are you more worried about your phone number getting hacked or your SSN getting hacked? If somebody can hack your nobody, nobody, you know, there's no impact. But phone number has a big impact. So if you connect the dots, your phone number and your email address, I wouldn't cancel out the email address. So either your email address or your phone number. They both are kind of your identity. They're kind of your national ID, so to speak, right? And and you know, in many countries, like let's just take India, for example, they have a national ID called Aadar, but you cannot use it until you get an OTP on the phone number that's connected to Aadar, right?

SPEAKER_01: You go to SyncPass, it's it's exactly the same thing. They connect you on your device. Uh you go to many other like another another country like Brazil, CPF, the same thing in our Argentina, the same thing in in Mexico, where you have claw ID. So, you know, uh uh Club Cleb CLEB ID, CLEB I think that's the right word. Uh and they have a KERP ID as well. But more importantly, right, your device kind of becomes your main access to I don't think yeah, your device kind of becomes your main access to it.

SPEAKER_01: So when you when you go with that notion in mind, what we are doing is connecting the device to your identity. So in many ways, if I see the world, all you have to do is be able to provide your phone number or your email address and be able to open that email address or phone number on your phone and somehow be able to bind it, scan a QR code, whatever that might be. As long as you're able to do that, you will be protected, your data will be protected, and you will not need this the third layer of you know uh the password fatigue that we have all gone through it, right?

SPEAKER_01: So the the ideal mechanism in my mind will be some kind of a pass key either face or something that you associate, you associate your device and you're done. There is no other log on as face blue, log on, and none of that would be needed at any given point of time. Because you're going from you know a hundred odd places where you have to open an account to like one or two that you might use for your day-to-day interactions.

SPEAKER_00: Okay. So so in in that case, like the people who are selling stuff where they need people to authenticate this way to allow them to purchase, like, do they pay you a licensing fee or how yeah, it will pay us the license fee, they will pay us the transaction fee every every time they verify you, verify your yeah. So you're you're you are kind of like like the uh credit card processing system, uh, you know, for the identity.

SPEAKER_00: Yeah, yeah. Okay, okay, okay. So that's what I wanted to understand. Okay, so uh what what is the financial model? Where is the money coming from? Okay. Okay. I I I understand, you know, um in the it's a licensing model. Okay, so you know, let's uh expand uh further. Now that you have this, you know, uh idea that you're working on that you're kind of, you know, it's a very nascent idea that you're trying to take uh mainstream. Uh while you are doing this, right, the core technology, let's say if it works, like you know, you might be the most dominant player here, uh, but you know, the the other uh you know uh larger companies might also come into this space.

SPEAKER_00: So using this core idea, like what are some other services that you could create that you know that would allow you to have a diversified portfolio that in case like this doesn't work, you you this same technology and what you have created is applicable in other places where that you can monetize.

SPEAKER_01: 100%, right? So Arna, so one of the things that is happening is that we are on a journey from status quo to the agentic journey, right? As we know, and as we know, it's gonna take some time. 100% it's gonna take some time for you and I or everybody to trust, to build that trust in this economy. So that's that's a given, right? Imagine when credit card came in. Not everybody wanted a credit card. They still wanted to use cash, it'll take some time, right? So what we are also so we are an AI first company.

SPEAKER_01: So that's one of the main things. Uh, we are connected to you know 2,000 plus data sources across the globe, a bunch of telcos across the globe. Uh we are also uh connected to a lot of financial institutions so we can verify your bank account, your debit card, and credit card across the globe, right? Uh what is happening is that we are kind of eating our ice cream, so to speak. So we have now kind of also created what we call as agent tech onboarding. So uh for a particular bank right now, I'm not sure when they're gonna participate in the agent league economy or whenever they feel, oh, an agent can come and create a bank account for you.

SPEAKER_01: Until then, they still have to do the verification, they still have to do the KYC, KYB, you know, due diligence. So they that's the business also that we are in. Those are our existing customers, you know, some of the large uh some of the tier two banks, uh, you know, some of the community banks that we're working with, they use our agent tick protocol to use agents to do agent tech onboarding. So, for example, you want to open a bank account and all you do is provide your phone number. We can verify your phone number or email address, we can reverse look up you.

SPEAKER_01: This is Arunasu, and then this is where he lives, this is what his date of brother is, this is what his phone number is, sorry, this is what his SSN is, and then you know, you don't even need to upload your uh driver license because we've given all that information back to the bank, and they can do an IL level one plus is what I call it, before they let you open the bank account, and then you know your transaction is completed. Now, the reason that they are able to do that is that we have built that trust, right?

SPEAKER_01: So we can deterministically say this email belongs to you, this phone number belongs to you, this identity is yours, and you are behind the screens. We build an amazing system to do that, and that's the that was the step one, Arunasu, before we get into agent tech layer, right? So that was the step one. But to even to build that, we build an agentic platform so that, you know, when you want to verify Ravi, you also have to verify Ravi from the point of view that, you know, is Ravi involved in any AML or is he part of any global AML and OFAC list?

SPEAKER_01: That's another part that banks and financial institutions and they all do that, right? The other one that we've also done is for non, you can say non-regulated use case where they all they want is, you know, just give me the name. I don't really care for their address. So we figured that out as well. But to prevent from synthetic IDs, we are also looking at your digital footprint. So by either by your email address or your phone number, I can tell. Do you have a LinkedIn account? Do you have a Google account?

SPEAKER_01: Do you have an Apple account? Do you have a, you know, all the digital footprint to say you are a real person? How long have you had this email address? What's your tenure of your phone number? You know, when was the last SimSwap Apple? When was the last quoting apple? All of those signals around we are connecting now, giving you, you can call it identity 360 or you know, consumer 360 to any financial institutions at the same time to any e-commerce giants to be able to serve their customers in a most frictionless way.

SPEAKER_00: Right? Yeah, yeah. I I think I think that that naturally brings me uh to our next section, which is becoming a platform. Like I did multiple work on this AML area, anti-money laundering, where you know I kind of built and rebuilt that platform. How to do the different levels, KYC, uh, you know, all the process that you're talking about. How how would you know different people at different levels can use the platform to do um you know uh AML risk assessment, right? Uh so your platform can actually do that AML risk assessment even for agents, right?

SPEAKER_00: So are you thinking about creating that platform where you know people can uh bring in multiple different uh transactions, multiple different data sources and multiple identities, and kind of plug it into your platform to create that risk scoring uh and kind of assess uh you know um is is this a risky transaction?

SPEAKER_01: 100% Arunas. You said it better than I would have. Yes, absolutely. You know, and and you know, think about it uh, you know, earlier the notion was microservices, right? At that point, you're like, oh my god, it's gonna be an integration nightmare. Right now, all we have is our MCP server, and you just, you know, our APIs that people are calling, in fact, we also are routed to our MCP server in a plain text English, maybe more better prompt, and then the response is send pack. But that MCP server response could be that this business is getting onboarded.

SPEAKER_01: I want to do an enterprise due diligence, but I also want you to monitor this particular business. Because, you know, for if you're an acquisition guy that's giving them a point of sale device, you want to monitor them throughout the year. Or just say I've onboarded, you know, uh Kickstart on August 20th. Based on the policies, every quarter I have to run OFAC and AML check. Keep this agent or connect this identity to your AML agent and make sure this identity is monitored. And at that point, it should be in that report.

SPEAKER_01: Earlier it was a nightmare. You would connect systems, you will buy more monitoring platforms. Today, an agent can do it. You know, of course, it's not as simple as SEP, but an agent can do it, connect it, put it in the registry, and then registry can be, you know, can be monitored, can be passed on in many ways, as you're saying. Yeah.

SPEAKER_00: Yeah, yeah, yeah. That's great. You know, uh the the there is a great potential for it to become uh that platform. Now that let's go to the next level um i about your team structure, right? How how how many people do you have in your company now?

SPEAKER_01: Uh I I actually, whenever somebody asks me their questions, I say hundreds and hundreds, right? So we consider our agents as our team as well. But uh overall, our team size is 21. Uh on an average, each and every team member is about five agents. So we are, you know, we start with 5x, you know, kind of, you know, expected is a 5x return in terms of you know making sure that the efficacy is there. And you know, uh the the best part out of this, uh, is that if we can run our company, you know, the way the way we want to run, you know, uh the way we want to run our you know customers' agents, we are kind of eating our own ice cream because we are seeing their live run in front of us.

SPEAKER_01: And more the more we do, the more we learn and the more better uh we make the agency clear as well.

SPEAKER_00: Yeah, yeah. So uh and and the people that you hired, right? Uh what level of freedom do they have in terms of creating new product or resetting existing product and like what level of ownership do they have that like you know, do they have equity or uh what would kind of make them want to think outside the box and think about larger picture for the company instead of you know just focusing on their robot?

SPEAKER_01: Yeah, so you know, one of the main criteria, Zarunasu, that we have in terms of hiring is this having this founder's mentality, right? It's it's not about, oh, you know, I'm I'm a DevOps guy and I'm done, right? Like, oh I but you know, if you come in our organization and you can ask a DevOps guy, hey, what is the use case about that customer? They will precisely tell you what that use case is because you know, we since we have a manageable team, we meet every day, we discuss all the problems together, and if somebody doesn't have to contribute, they can sit there and learn, and then they can connect the dots on on the main task that they have, and then the in the and what I call this as the founder's task that they have, right?

SPEAKER_01: And and we are we are uh you know kind of we're trying to do something which is very normal, which is we're trying to keep organization very flat. Yes, there will, there's definitely layers of accountability, but in terms of, you know, we don't believe in ranks. We, you know, uh anybody can be wrong, any and everybody can be right in terms of as we are progressing. So our really thought process, the way it works is customers, people, and then revenue. Because if customers will be happy, people are gonna enjoy working with them and revenue will come.

SPEAKER_01: That's there is no doubt in my mind. We this is my third startup, so it's it's it's easier, you know, and I have the scar tissues to prove what all I learned uh at some point, right? But what we also believe is also customer-driven innovation, right? So, you know, it's yeah, if you remember the whole blockchain thing that we learned from, right? It was a solution looking for a problem. I don't know how many times, you know, I heard, oh, we're gonna use it for this, we're gonna, and finally now people are actually catching on and using it for whatever right capability it is, right?

SPEAKER_01: So our mantra is have a founder's mentality, but also have customer-driven innovation. Most of the products that we have, Arunasu is has an active customer, right? And and you know, or or has a larger potential pipeline. So one example that I'll give you is that we were work again, identity is the key element, right? Uh, but one of the innovations that we worked on is that I can figure out how many vehicles you have during onboarding. And you know, the reason we've been there is that, okay, now the insurance industry would absolutely need that, right?

SPEAKER_01: Imagine uh you going to any platform and like I want to apply for you know insurance, and then all you do is provide your phone number and everything is done, and you just pick the vehicles that you want, it makes it very, very seamless, right? And then, you know, so so as much of like uh uh like we think it's innovative, it was driven by a customer use case and a pipeline, not really sign, but uh and a pipeline, and and you know, we were able to accomplish that. So the the the short answer, uh I give you a very long one.

SPEAKER_01: The short answer is that I have a founder's mentality, but it has to correlate with customer-driven innovation.

SPEAKER_00: Yeah, yeah, that that that's that's a good philosophy. Uh let's uh you know talk or talk about the final stage. Now, let's say you know um it all goes well and you become the dominant player in this space, right? Where do you see, you know, this is still like a very early uh level development, right? Uh it's a concept that people uh have got thought about. Uh it's it's it's a service that you know people need it, they just don't know it yet, right? They don't know what it is and and you know uh uh when they would use it, but it's it's a matter of uh you know um kind of wrapping their head around it, right?

SPEAKER_00: So uh, you know, this is already kind of a very innovative way, very uh future uh thinking uh you know project that you are working on. But if you had unlimited funds to invest into research and development, what are some fields around this where you think there is more research needed that can be really game-changing and it's really essential?

SPEAKER_01: That's that's a that's a very, very good question, Arunas. Uh so you know, uh there are the way I think about is that one is what's the ground level thinking and like you said, what's the end vision, right? Uh so the end vision is eliminate fraud, right? That that's what that's the main purpose of building a regulated product and so on and so forth, right? That's the end vision. But along the journey, uh, you know, some of the key things that we are gonna that we want to work with our customers is that we also want to be able to in many sh in many ways uh be able to give you know customers, uh customers fraud flags back to them, right?

SPEAKER_01: So for example, if Ravi is an identity and I did some fraud somewhere, and you know, there is a there is a notion to that call first party, third party, you know, fourth party fraud, whatever party that's there, right? But in many ways, what we truly believe is you know there has to be a golden record for identity, right? That's that's the term that we all use in our MDM space, right? And then there is there is a version uh of the record that that is being used across the internet. Wherever you go, you create an account with limited information, right?

SPEAKER_01: What we want to do is be able to connect the two and then have that linkage always be there, right? And now I I really don't have a very specific answer to that, right? But one of the main things that our goal is that if I get unlimited funds, we want to manage the identity across the globe, right? In terms of being able to manage more than you know, six, seven, eight billion identities, so that you know, most of the interaction becomes frictionless. Most of you know, the the pro the other larger problem that will get solved would be financial inclusion.

SPEAKER_01: A lot of you know, folks don't get access to the financial products that they're not able to verify themselves. So that's one other thing uh that we would like to tackle, right? At some point, all you need is your phone number and your selfie. I think the rest should be accounted for. You should not be able to carry your passport anywhere. I get it, you know, that's there, but that's my you know, like like true, true uh unicorn vision, so to speak, right? Yeah, yeah. I mean, yeah. Very close. Yeah.

SPEAKER_00: Yeah, yeah. I mean, identity itself is is an ever-evolving uh research field, and uh you know, adding AI to it really makes it um uh very futuristic. So it it it it's a field where a lot of research is needed. So thank you for all the work that you are doing. Now let's talk about the uh you know managing the external environment. On one hand, you're you're kind of working on a new standard that does not exist yet, yeah. Right? So on one side, like you know, what are you doing to have a say in creating that you know open standard for agetic e-commerce?

SPEAKER_00: On the other side, what are you doing, you know, uh, to your future customers, the banks or uh the e-commerce players, to you know, help them understand that they need this.

SPEAKER_01: Yeah. So uh thanks again. That's again a uh uh a very involved question, so to speak, around usually. So one thing is that we are part of the ecosystem. So we're working inside an ecosystem. So first question that you said, what are we doing? We're building the right set of partnerships around usually. So we are we are working with you know the the the right parties, so to speak. The the the bureaus, the payment gateways, the CDNs, uh the pay uh, you know, the the the commercial entities or or the uh or the you know e-commerce giants that are providing services to you and me in many ways, right?

SPEAKER_01: Uh and then so we're building partnerships, so to speak, right? And then we are also working together to speak the same language. So we all want to say, hey, when uh when you know Kickstart is sending that token, I can add to that token, but that token has been verified, and we all have to kind of shake hands to say that's you know uh uh that's an agreed upon format, that's an agreed upon you know, uh attributes that have to go. So that that's agreement is is also out there. Uh the the second part is that it it's it's the it's the uh you know it's the Economy and then it's it's the commerce doing it to itself, right?

SPEAKER_01: So if more consumers start going on to the the uh the LLMs and everywhere else, and the the transaction starts from there, the the businesses have to come in, they have to play out there, right? They they have to go meet the customers where they are, so to speak. So it's doing its own job, and then you know, uh the the media is is helping a lot, right? So like if you think about it, if you go and tell any I I I can tell you this, you can go and talk to any elderly person probably 10 years back and say, oh, you know, there is something called identity verification, like yeah, you know, that's that's bullshit.

SPEAKER_01: Now they will know what is an identity fraud, they will know what is a breach, they will know what informations of their in uh of their identity that can be breached, that can hurt them. So people are more educated, so to speak, right? But you know, unfortunately, I'm not sure there is still a lot of fraud going on, there is still a romance fraud that I heard the other day that some uh some elderly person was giving money to Dolly Pardon. Like, think about it, she's no she's not even alive now. But imagine somebody giving money to such a large so you know there are still some those kind of uh frauds that are going on, and you know, it doesn't need a complete solution, it needs that patchwork of solution to fix it today, because uh the wave, and any wave, that's what happened, right?

SPEAKER_01: Even when we saw the digital economy came in, the larger banks were first to jump onto that, if you if you remember that, right? Like Bank of America, I I don't know if that one of those banks had the first app out there, digital banking app, and then this everybody followed the suite. So that in many ways could be a traditional way, but I feel like we are just uh time in waiting for a clawbot moment again. You know, if you remember that, you know, about 10 million people bought uh installed clawbot in a day, or I think it was more than 10 million.

SPEAKER_01: I'm just being very conservative here. But we are waiting for that clawbot moment for it to happen, and either it can happen this Thanksgiving, whenever the shopping is at the peak, you know, that's when people get accustomed to this, or it will happen at some point uh pretty soon. But the clawbot moment is coming.

SPEAKER_00: Yeah, definitely. So uh that brings us uh to the end of the episode. Okay. Uh if this episode shifted how you think about AI in your organization, send it to one person who needs to hear it. And if you're ready to act, visit tipswora.com to get certified or reach me directly at arrangsepatnite.com. The intelligence economy is already here. The question is whether you are building for it. That's a wrap on today's episode of What Comes Next. If this conversation gave you a new way to think about AI strategy, share it with one person who needs it.

SPEAKER_00: You can find everything I am building at tipsura.com, including AI certifications for you and your team. Connect with me on LinkedIn by looking up my name, Arnan Chipat Narik. You can find Ravi on LinkedIn or visit his company. I would like to clarify for our listeners. Uh, I said KYX start, it's pronounced Kickstart AI. Uh if you are a startup, you can sign up for an account and you get 100 transactions for free. Okay. Thank you, Rabi, for uh sharing all this information with us. It was been great speaking to you.

SPEAKER_00: Uh amazing IPO or uh, whatever you are looking for in the next round of funding. So good luck with uh all your future and um thanks for everything you are doing.

SPEAKER_01: Appreciate it. Thanks, Arunasu, for having us. You know, we just want to be of service. You know, we we are very excited, and uh hopefully we can uh we can be a very critical component of agent e-commerce.

SPEAKER_00: Okay. Until next time, build the architecture, and the advantage follows.

The Story