Weekly AI Update with Pavan Agarwal: Chat Is the Future: How AI and Blockchain Are Reshaping Global Banking

Weekly AI Update with Pavan Agarwal: Chat Is the Future: How AI and Blockchain Are Reshaping Global Banking

In this episode Weekly AI Update, David Lykken sits down with visionary innovator Pavan Agarwal, CEO of Angel AI, to discuss how chat-based artificial intelligence and blockchain technology are revolutionizing global banking and lending. Fresh off his meetings with top financial leaders and government officials in Saudi Arabia and at the North American Blockchain Summit, Pavan shares how nations abroad are rapidly adopting AI-driven, chat-first banking systems—while many U.S. institutions risk falling behind. From achieving zero defects in federal audits to envisioning a 2030 world of 24/7 blockchain-powered homebuying, this conversation reveals why the future of finance is conversational, compliant, and borderless.

[David] Listeners, it’s time for another AI update, and we’ve got Pavan back with some really interesting information. But folks, this guy is such a world traveler. We gotta get caught up first on where have you been and what have you been doing? Saudi Arabia, Riyadh, Pavan.

[Pavan] Yeah, so just got back from Saudi Arabia, then I’m after that I went to Dallas. I’m here. So, Saudi Arabia got home for two days, Dallas for the North American Blockchain Summit. It just, it’s just exciting. The world is changing so fast and if you’re not up to date on Blockchain, Crypto, AI, you’re gonna miss the boat. And where this stuff is happening, right? It’s not, I’m sorry to say it’s not happening at the MBA convention. I mean, there’s the future is not at the MBA convention. No, no, it’s not a slam on the MBA.

[David] No, there’s so much happening, but and it’s really significant. Talk about what the Saudis had you over there for because that was really significant. They’re seeing the value of this. And I think how many things do we start in a country and then another country sees the value in it, they jump on it before we here at home jump on it.

[Pavan] Yes, it’s chat is an example of invented in America but commercialized and as you know, abroad. So, you know, last time we talked about the conversation that Eric Schmidt, the founder of Google, had where it said that traditional user interfaces, Windows user interfaces, are dead. Well, the rest of the world figured that had figured that out over ten years ago. WeChat runs all of China, okay, and you tap your phone and all your payments are done. So if you don’t have a WeChat account and everything you want done, you send a quick message and it’s done. And that’s the most convenient way. Short little text message, you send a WeChat, everything is done. But it’s only America where that’s not the case. It’s in China, it’s in India, it’s in Japan, it’s in North South Korea. The rest of the world had figured out that chat is the future of all commerce. And we’re talking about 2013, 2015, and WeChat had taken over China. And you tap your phone and you made all your payments. You tap your phone, you transferred, you transferred money from one account to the other account. Right? You send a little chat, you open up WeChat, send a little chat message, everything got done from brief two seconds. But nobody wants to pick up the phone and be on hold and wait for an agent. If you could do that with a quick message, and nobody wants to navigate through a long web page. You just want to say the most natural interface for human beings is to just say, Give me my loan approval. Right? Yeah, the most natural interface for every human being is to say, hey, fund my loan. Here’s my bank statements, here’s my pay stuff. That’s how human beings want to do it. You don’t want to like go click a hundred buttons, be on hold for 30 minutes or whatever, right? They just say, you know, what’s the price today? You know, what’s the difference between FHA and conventional? You know, what if I put more down payment? All right. It’s it should be your brain, what you desire should be driving the computer, not what the computer desires. So the traditional web interfaces, the Windows basic interfaces were driven by what the what was convenient to the developer of those platforms. But when you when you flip it around and you make it AI driven, right, the AI is now your servant. You’re the master, the AI is a servant. Right? Because the AI is ambidextrous, right? and it can do whatever you ask it to do. It’s not rigid. So it is it is driven on your satisfaction. So chat is the really the only way. It’s the only thing that’s going to exist in the future is a chat-based interface. And chat is in comes in many forms. It could be voice, it could be conversation, it could be video, like you know, but like we have with AI twin. There could be voice and video, but it’s still fundamentally under if you lift the cover, it’s all chat. It’s message, short messages going back and forth, which is right, and essentially in the real life.

[David] So it’s not surprising that that is working everywhere else, and it’s really only now catching on here in America. I mean, we say, well, the chat’s been around for a while. Yes, it has, but not to the degree that where you can do all the transactions. That’s the part, it’s the transactional components of chat overseas with the WeChat.

[Pavan] Correct. Correct, right, right. It’s pretty much in every other country except America. And the problem, the problem isn’t Americans. Americans want it also, especially anyone that’s 30 and younger, they definitely want it because they hate being on the phone. All right, they just want to send a quick chat message and be done with, right? And the median home age, the median home buyer age is something like 30 or 35. So if you’re delivering, if you’re building applications that aren’t directed to that demographic, you’re gonna miss out. Yeah. Okay, so but that goes full circle as to why why Saudi Arabia invited us to um to Riyadh, is because the banks there understand that if they continue to do business in the traditional manner, the old tech manner, which is, you know, and I went to one of the banks and I and I walked in and it looked just like an American bank. We had a huge floor, rows and rows of cubicles of people on the phone and the computer clicking away. Right? And and and they’re like, and and this the the bank, top bank executive that looked at me and said, We don’t want this. We don’t want to repeat your mistakes. Right? Our customers, the Saudis, are demanding to just do it over chat. They don’t want to have to go through the hassle of having to explain something over the phone to a human being. And especially when it comes to finance, because nobody wants the embarrassment and having to deal with like another human being saying, Oh, is that all you make? Or, you know, where do you work? Right? The the whole thing about you know potentially disapproving of your lifestyle. Right? Whereas if you’re chatting, that is eliminated, especially if you’re chatting with an AI. So the rest of the world’s forgotten, they’re like, they’re like, they literally said, I don’t care what it takes, we want angel AI in Saudi Arabia. We want it for our bank, right? You just get it implemented because our customers, like, we want to be the first bank and the only bank in Saudi Arabia to have this end-to-end solution that’s completely AI and we’re gonna dominate the market. The banks I was talking to, they already have like 70% of the of the mortgage market. They want to make sure this stays that way. so you know, that’s what’s going on in Saudi Arabia. And it’s not just Saudi Arabia, it’s it’s we met with the in Riyadh, we met with the top banking leaders and investment houses and from Qatar and also from the UAE. They’re saying the same thing. Like this is this is the future. Right? And this is what we want. We don’t want to be stuck in the past. There, they like there’s this. I think American businesses have got gotten complacent, and like I was saying, it’s not American, it’s not the fault of Americans. They want it as well. But American businesses, right, they’re so used to being on the top of the food chain, think they can just not have to be the most competitive. They’re forgetting to put the customer first, right? And they’re not paying attention to what the customer is asking them for. The customer wants the ease and simplicity of hey, two seconds, then send a quick command, and then and then my request is fulfilled. That’s what a customer wants. But that means a full reinvention of who they are, right? Whereas other countries, it’s a blank slate with them, right? It’s starting from, essentially starting from especially Saudi Arabia. It just opened up the country, and effective January for the first time ever, non-Saudis can buy real estate anywhere inside, just about anywhere in Saudi Arabia. Right? So complete open. They’re going from completely closed to completely open and like at like warped speed. And that’s why they had the comedy festival. I mean, it’s unheard of to have a comedy festival in Saudi Arabia where you know, complete open conversation, and you can make it and you’re you’re allowed to make fun of anything. Including the royalty, including the royal family, you know. So, I mean, Russell Peters was there, who’s by the way, one of our partners, a great spokesperson, and we went to his show, and he was there, and you had the royal family right there in the front row, and he was making fun of the royal family, right? And it and they were laughing, they were enjoying, they were enjoying the process, right? It’s it’s like you know, we we it it’s it’s good leadership when you can make fun of yourself when you when you can when you can take things in strike.

[David]: That’s amazing. Yeah, that’s an amazing story. But I would remind our listeners that are listening to this that at the core of Angel AI, it’s based on chat. And you continue to enhance it and enhance it and work on that. And a lot of people say, Yeah, but it’s not this data entry like this. No, that’s not what the that’s what the next generation wants. Who’s buying homes of the next generation? Yeah, it’s the ones that want a chat. You guys right. One of the things I want to talk about in this podcast today is the regulatory environment where billions of dollars have been assessed to institutions because of errors. Jamie Diamond once said, mortgage business is a lousy business. I don’t want to be in it because, in part because of the penalties, the regulatory more you make if you have a good year, it’s like the DOJ, at least under the Obama administration, was going back. Well, that’s a great place to start taxing them through these regulations. But you went through an experience at the same time. This started, I think it’s I’m looking at your report that you sent over to me. It’s a list of all the money that was paid out from I think it was 2015 all the way through 2019. I mean, it’s multiple billions and billions of dollars. They went after everybody. It was indiscriminate. They went across the landscape looking for errors, what was going on, and you had an acceptable where everyone else is getting penalized in the millions to the hundreds of millions to the billions. You got a letter of commendation of what an amazing job you’re doing. Talk about why that was and what you saw. And this is important, listeners, because we’re and everyone thinks well we got the Trump administration, they’re going easy in regulation. Don’t be l living in that. We’re we still have a regulatory body, and they’re gonna be coming after those that are not doing it right. And that’s why it’s so important that we be compliant and doing it now with chat, but being compliant, but those that don’t are gonna be paying a price. Pavan, what did you learn out of that 2015 to 2019 period with your technology?

[Pavan] Yeah, first thing we didn’t get a letter of accommodation. The office of the inspector general does not do that. But they but they didn’t have to be able to do that.

[David] I didn’t mean to say a letter, right? But anyway, you got it. You they they came back and said, Yeah, we did just they they just quietly went away.

[Pavan] They said, okay, you’re fine. They just they don’t they don’t say anything. They just No fine, no fine. Right. So you know, there’s something called the False Claims Act that was there’s an old, like a 150-year-old act that was that was written during the Civil War, uh, you know, to find people who were cheating the U.S. government during wartime uh for you know military sales to the U.S. government. So uh the they got smart about you applying that act to other parts of the economy um in over the last 20 years, right? They went after defense companies, and then they went after the banking industry. And the way they went after the banking industry, well, especially the mortgage lending industry, the way they went after the mortgage lending industry was they would they you would they’d issue your subpoena and they would ask you for a report of all the production you’ve ever done or over a certain time period, and then and then they would randomly select roughly 100 files. Uh in our case, they selected 200 files, 100 100 forward and 100 reverse, 100 forward mortgages and 100 reverse mortgages, heck of home equity conversion mortgages. So 200 files they selected. And what they normally do is they take this random sample uh across a wide range of years, and they have then these forensic auditors like go through it, like a whole massive team of forensic artists go through it document by document to find something that was done that didn’t meet the guidelines, that didn’t meet the FHA ranks. Okay, and if they find a document that’s out of place, then they mark that file as defective. Okay, so what was happening is like all those you know that list there’s so many people that were hit that was from fines from as much as a billion dollars to $30 million, $20 million, $50 million.

[David] And so what they would do is it’s like $50 million to Deutsche Bank, that’s staggering.

[Pavan] Yeah, yeah. It’s incredible. Yeah. So what what they would do then is they said, okay, we let’s say we audited a hundred files and we found 75% were defected. Okay, and then they’ll say, okay, well, 75% of the files is you have a defect rate of 75%. That means 75% of the claims that we’ve ever paid out never should have been paid out. Okay, and then so they take a look at the claims they paid out that that they’ve that you that they paid out and what they expect to pay out in the future, and they take 75% of that dollar amount, and then they put there’s treble damages in that act, and then times three, right, and then they give you a bill, yeah, that’s potentially in the billions of dollars, and say, okay, you all you owe the US government this much money. So that’s the quick skinny on the false claim act and how it works. Okay, so and that’s why those penalties were so massive, and you know, they collected so much money. Okay, so you know that so they were sending these subpoenas out across the industry to basically you know every major lender, and we obviously are major lenders and however, in our case, and I think we’re the only ones that had this where they audited 200 loans, 100 reverse and 100 forward, and it was the time period was like from 2008 to like 2014. Okay, and they were astonished because there was zero defects.

[David] And that was because

[Pavan] So, right? So it was like literally zero, so zero times. So that means that there was there was that was their process, zero defects, though I we don’t owe them any money, and they went away. Right? So you know you don’t you you don’t get a pat in the back, you don’t say they don’t they don’t even say good job, they just they just go away.

[David] They write you a letter just go away as a commendation as far as I’m concerned. So yeah, I’m gonna stick to my it was a comment.

[Pavan] Yeah. No commendation. They just they just say thank you very much, thank you for your time. And they you know they go on to the next slide. Uh well, yeah, because I mean that’s it’s uh it’s a subpoena for the federal government. So you get a whole army of uh of attorneys. I mean, they beyond just auditing the files, they interview they interview almost every employee, they subpoena all the emails or read through everything. It’s an extensive, extensive uh investigation.

[David] They walked away without finding you anything, and you’re one of the few that went unscathed. Amazing.

[Pavan] I don’t know of anyone else who’s gone through that and had zero defense. I don’t think it’s possible. And you know, we had conversations with that.

[David] Is that because your underwriting is so conservative and you guys are so tight? That I mean I know it’s a rhetorical question. The answer is I used to sell loans for you guys. I know that you guys were reasonable in your underwriting.

[Pavan] Yeah, no, I actually we were doing you know, it’s in some of those years we were doing like two billion, three billion of production a month and there are there was also the reverse mortgages covered the biggest years. We’re the years that we did the biggest amounts of reverse production, like over 100 million per month reverse production, which is huge for reverse and so this covered some of our biggest production months, and that’s what they were after, because they they’d only come after you if you were doing a lot of production. Because they figured there’s a lot of money there. And the reason they didn’t find anything, because you know, although Angel AI as a brand, as a complete end-to-end product, went into full production. You know, end-to-end meaning from the chat interface that we just talked about to where you could run the whole loan origination process right from the chat interface where consumer can talk from the chat all the way through closing, right? That complete end-to-end connection was put into production in 2019. Okay, however, the core engine, the core underwriting engine, the AI underwriting the loan files, that has existed for years, right? So, I mean, the loans that we did in 2008, 2006, 2007, they were all underwritten. All those heck of loans, all those reverse mortgages, and most of those forward mortgages from that time period were all underwritten by the AI. Okay. And so what is what does the AI do? The AI has digested all of the guidelines. Okay, right, it’s trained on the on the entire 4001 or back then it was 4155, right? And it’s trained on all the letters, it’s trained on all the Ginne Mae guidelines, trained on everything. It’s all there. Right? So when a document comes in, it just you know it can’t get it wrong because it’s a computer doing it. It’s not a human. Humans, computers, you know, if it’s trained properly, it’s not going to make those kinds of mistakes. All right, and so you know, every single one of our appraisals were underwritten perfectly, right? It it’s not a matter of whether it’s conservative or not conservative, it’s just it’s a matter of is it right or not right? Okay, and because the the guidelines are black and white. Is that is it there or is it out there? Right? And so it was is that it’s that simple, right? And so the tech works, right? And it’s been in place for years. And you know, we talk about we’ve done 35 billion since 2019 on Angel AI. And that’s 29, that’s 35 billion since 2019, uh, where it was end-to-end Angel AI. But if I actually go back to 2006, right, and it was in 2006 where we sent out a press release that we announced the world’s first automated underwriting reverse mortgage engine. Okay, it was literally, you know, our technology was the first reverse mortgage automated underwriting engine. That was back in 2006. And it’s because of that automated underwriting that happened, the reverse mortgage underwrite completely by the machine, right, that they didn’t find uh, you know, the government when they did the audit and the investigation, not a single loan they found that was underwritten incorrectly. It’s amazing. It’s just it’s the difference between you know you know producing an automobile with robots or producing it by hand, right? If you produce it with robots, it’s you’re going to have a higher quality automobile. It’s not going to have it’s not going to have production line default.

[David] So you have continued to do that. I think the reason I wanted to bring this up, listeners, is because as we’re facing you know, it might be a little bit easier in the regulatory environment we’re at, but we’re still dealing with a highly complex transaction with lots of areas where something can break because of the human element. We have got to be using AI more and more integrated wholly and completely into our manufacturing process. Pavan, kudos to you and your success in building an amazing technology and what you’re doing. Now you’re at blockchain, and now you’re talking where does this all go with blockchain? Let’s fast forward into what the world looks like in a mortgage with blockchain.

[Pavan] I’d say let’s go through a time machine. All right, we’re going through a time machine, we’re falling down this time machine, and it’s 2030 now, five years later. Right? And AI and blockchain is ubiquitous, it’s everywhere. Everyone has an angel wallet installed, everyone has angel AI installed on their phone, everyone’s transacting with stable coins, and in your wallet, you have in your in your decentralized blockchain wallet, you have multiple assets in there. You have US dollars, you have you have your securities, you have Bitcoin, you have Ethereum, you have gold, right? So you’re managing all you’re your own central bank. And it’s all you’re controlling all that in in your wallet. Okay, and then you have your AI on top of it that’s helping you optimize the so that you get the highest possible return, highest yield on the assets you have. You have your assets and your debts all in one wallet, and the AI is optimizing it for the best return. Right? That means maybe moving some money from dollars to gold or gold to equities, or moving into Bitcoin or out of Bitcoin or whatever it is, that it’s gonna find the lowest, the highest return for you possible, right? There’s so many people who have a mountain of credit card debt and they’re paying 18% credit card and they have fifty thousand dollars in the bank that they’re earning 0.1%. Right, so you have $50,000 only 0.1% in a in an environment where the tenure treasury is that is at four and a quarter. Number one, that’s crazy. You should be earning four and a quarter on that, on that fifty thousand dollars. Number two, why don’t you take that take the fifty thousand dollars and pay off your 18% credit card debt? Yep. Right? And that’s how you start saving money. And like most people either don’t know or don’t understand these things, or no, that’s number one. Number two, don’t have the time to think about it. And they don’t they don’t know how to regulate. Yeah, right. I mean, they’re just like there’s so many things going on in life. Life is hard, and you don’t have time to think about these things. Well, if the AI think about it, you know, it’s the AI is like take the thousand top economists of the world, take all of their intelligence, put it in one place, and that’s the AI. So you can’t get anything smarter than the AI because it’s the collective intelligence of the best, best economists, best traders. Right? And not only that, but it’s got it’s a machine, so it can it could in a in a microsecond run all the scenarios, all the trading options, right, and all the predictive trading options for you, and then say this is the best position for you to have. Okay, so that’s what 2020 or 2030 looks like. And it all what it also means is with stablecoin being ubiquitous, that means you could fund your loan. You could you can buy your house 24-7. You don’t have to wait for wires to get cleared. You literally can send stable coins across and purchase your home. There’s no more, there’s no need for a closing agent because a blockchain smart contract is controlling the escrow deposit, the repair, repair escrow, all of these things, the seller and the buyer disbursement, the commission disbursements, it’s all being controlled by a single smart contract on chain. So you can literally fund your transaction uh 24-7. The cost of title insurance goes down to a fraction of what it is because the risk is gone. The largest risk in the title insurance process is the closing protection letter. And if a smart contract is controlling the closing and your money’s all held on chain, right, there’s no there’s no need for a closing protection letter. And that’s 90% of the risk a title company carries. So that’s eliminated.

[David] Yeah, that’s a great point.

[Pavan] At the at the end of the day, a hundred and eighty billion dollars a year in economics and in cost is sucked out of the economy every year because of the lack of blockchain technology because people don’t trust each other, right? And you have to go through all these intermediaries who take large fees along the way, right? So you’re talking about 180 billion to 200 billion dollars annually is lost out of the pockets of hardworking Americans, right, and it’s being lost into this ether and where it should be going back into the pockets of consumers. Think of how much more efficient we become as a country, right? When that money is put back into people’s pockets so they can go back into you know consuming, paying for education, paying for their medical uh needs and so forth, right?

[David] That’s a great point. The future is coming on us so fast, and the way you’ve built this system, for those that are not moving towards this or involved in this, are they already dead? Do we have the walking dead scenario going?

[Pavan] Well, I don’t know if I should go down that path, but I would say you can’t ignore it. Right, right. You can’t I mean you can ignore it. I mean you can ignore it. You could probably get away for another year or two of ignoring it, but it’s it. Right, and you know, customers, look, customers want two things, right? They want convenience and value. And they’re gonna hit the easy button every single time. Whatever button gives them the most convenience, the most value is what they’re gonna hit. Guess what? 24-7 transactions, 24-7 answers, perfect products being produced robotically, right? Um answers when they want them, answers they could rely on, right? no need to worry about uh you know conflicts of interest, right? no embarrassment by talking to bankers who may or may not eat and when and you don’t know if your banker has your best interest at heart or not, right? so you eliminate all of this inconvenient, all this cost from the process. Guess what? What are consumers going to do? They’re going to hit the button that’s greatest value and most convenient. Right, they’re going to hit the easy button. Just like you know, we saw this in the mortgage industry in the 90s, where people were saying, Oh, you know, no one’s going to ever do a mortgage on the internet. Yeah, where did that end up? I was there, I was like, you know, we I used to go to the MBA all the time, and it’s like 90% of people would say, that’s a nice little fancy idea, but no one’s going to ever do mortgages on the internet. I call BS on that. And you did, and look at the internet. And today it’s the same thing. No one no one’s going to do the common thing is no one’s going to ever do a mortgage by AI, entirely by AI. No one’s going to chat for their mortgage. No. They will. Guess what? The rest of the world, that’s there. The difference is that the industries are entrenched to the US and they’re trying to set the agenda that they’re trying to tell the consumers, no, you don’t need value, you don’t need convenience. You need to pay us because you don’t need value and convenience. That’s the mindset in of US companies right now. It’s just sort of a self-preservation with a dash of arrogance. Right. Whereas the rest of the world, the rest of the world is so hungry, right? So hungry to say we could do it better. The rest of the world is so hungry to catch up and exceed the US that they’re driven to find better solutions. Their companies are driven to find better solutions, and the people are driven to find better solutions. And saying, look, we’re going to rise and we’re going to do it better than America is doing. We love America, we love the culture, we love everything about Americans. And the kind of interesting thing is like if you’re an American in the Middle East and you have and you speak like an American, you get a lot of respect because they just love America so much. But at the same time, at the same time, they’re so because they look up to America, they’re so motivated to show they can do it better than America. And so they got a fire that we could we’re going to show that we can do better than America just out of national pride. So we need some of that national pride back in America saying, no, we’re going to stay number one. And we’re going to innovate and come up with better ideas than everyone else. Right? And we’re not going to stick with the status quo.

[David] And we’re doing and you’re doing that.

[Pavan] Otherwise, color TVs are gonna be taken over by Sony, right? Yeah, the trans remember the transistor was invented in America. And then they didn’t get the inventor of the transistor didn’t get any traction here. And Sony bought the idea. I how silly is that? Yeah. There’s so many stories. All right, the microchip, Texas, yeah, Texas Instrument was a giant in many in manufacturing uh microchips. Microchips were invented in America. But the who’s the greatest producer of microchips today is TSMC in Taiwan. Why do we let these things happen?

[David] You ought to run for office, that it’s starting to sound like a stump speech there with what you’re saying. But it’s really true. That’s actually true. And it’s been so good to have you back on talking about this stuff, there are so many moving pieces to this, but you are so far out ahead of everyone. It is really encouraging to see the example that you’re setting. It’s not too late for others to get involved. But today we covered the regulatory side of it, how expensive it can be when you do it wrong. And then here you are in the same environment, skate through, and they just packed up and says, Good job, you’re doing it. Keep up the good work and left and no penalties. That’s a pretty profound statement. Thanks for being here, PavAn. Enjoy the time. It was good watching social media. I’m seeing that you were with Governor Abbott yesterday. You work with world leaders, and now you’re we and you’re meeting here with our Texas governor. That’s pretty amazing. I love it.

[Pavan] Governor Abbott, he’s that’s a great guy. And he really cares, and he listens, he listens, he gives, you know. He listens because he’s smart. And in that room, though, there was about you know eight or ten of us, and it was a wide variety of people. And we talked everything from energy production to blockchain to you know, obviously lending practices that I was talking about to education, right? And he got like he was an expert in every area. Like he got it. Like, wow, right? The guy knows everything. Like that’s what that’s what makes him so smart.

[David] He’s an effective leader, yeah.

[Pavan] It’s like, how does he possibly know it? He knows all of it and he integrates it all together. The guy is just off the charts smart.

[David] Yeah, it was fun seeing you beating with our governor. Thanks for being here, PavAn. Have a great rest of your conference. We look forward to uh getting you back here next week and give us an update on the blockchain conference. Thank you, friend.

[Pavan]  Thank you. Cheers


Important Links

Pavan Agarwal is a renowned leader in the mortgage lending industry and a pioneer in bringing artificial intelligence to the financial markets. Agarwal serves as the President and CEO of Sun West Mortgage Company and Celligence International, LLC.