Rebuilding Cross-Border Payments with Prabhakar Reddy of OpenFX

0:05Prabhakar Reddy: Welcome to ReBank, your front row seat to the future of finance.

0:13Prabhakar Reddy: Prabhakar, welcome to ReBank. Thank you. Thank you for having me, Bill. Yeah, it's great to connect with you. I've been following your work at OpenFX for some time. The payment space FX specifically has been something we've covered on ReBank over the years, including with some of the maybe last phase or first phase fintech incumbents, the likes of TransferWise, CurrencyCloud, companies that have done an incredible job of progressing this space and really scaling volumes and ultimately underpinning a lot of what has powered, I think, fintech offerings over the past 10 years at this point. So it's exciting to connect with you and understand what your take is on this industry and how you see things evolving from here.

0:57Prabhakar Reddy: But firstly, welcome and thanks for making the time. Thank you. It's a pleasure to be here. Looking forward to having a fruitful discussion today. When did you start the business? All right. So the company was officially started in Jan of 2024. But the idea and the research was happening a few months prior to that. And let's talk about the origin story of OpenFX for a minute here. I actually did start this company with the intent of starting the company. You know, I actually went down this rabbit hole of cleaning various layers of the onion to understand how global cross-border money moment works.

1:31Originally, we ended up investing in a good company if someone's actually solving this problem. So a lot of things are happening parallelly in my life. So I grew up in Dubai and I spent, you know, my life through 80s, 90s, 2000s in Dubai. I moved to the U.S. in 2012, came back to Dubai in 2022. to, I realized I started noticing a lot of lines around Western Union and a lot of exchange houses. It was very reminiscent to what I saw in the 90s as well. So that was one data point. A second data point was seeing parallelly on TechCrunch. Lots of announcements that stablecoins have found PMF, cross-border money movement has been solved by platitude of companies basically that have been raising money from really strong VCs.

2:13Like, okay, there's a big disconnect here. That's number two. So number three, I was basically looking at figuring out what's next in my life. I'd taken a step back from my operating role at Falcon X, then I was trying to figure out whether I should be a VC, whether I should go back to being an operator. And just out of curiosity, I started building various layers of the sunning and met every single company in the space. I realized that after having met a lot of fintechs, the problem with cross-border money movement is not at the stablecoin layer. It's purely the FX layer because most of the infrastructure that's been built was built as a wrapper on banks.

2:47And banks are in the business of holding money, not moving money. And the minute you realize that, you realize, you know, there's a reason money doesn't move after 4 p.m. EST. Money doesn't move on weekends. Money doesn't move on national holidays. And it takes three to five days because of the intermediary banking infrastructure, the Nostro-Australian infrastructure that is set up. So it's like trying to peel layer after layer and realize that it's a very complex business to build. If you really want to make money movement happen, like sending the text message on the iOS network, where it's near instant 247, you'll have to build this infrastructure from ground up.

3:16And somebody who understands market making, somebody who understands stablecoins, somebody who understands FX, somebody who understands licensing, regulation, compliance, the ability to raise an inordinate amount of capital as well. I was like, okay, there are very few people in this world who can actually solve this problem and operate this entire business at a business model that actually makes sense, right? Because if you really want to make cross-border money moment happen, you'll instantly at extremely low cost. The business model itself requests transformation. So that is basically where the origins of OpenFX as a business came about to be.

3:46Spent a few months doing the research, started the company in Jan of 24. And we did our series A about a couple of months ago. Wow.

3:54Prabhakar Reddy: Well, congratulations. Amazing founding story. And it sounds like a very strong early trajectory. Not surprisingly, I guess, given the size of the problem. But maybe just starting from like FX 101. What is FX? How do you think about what it even is? That's a very good question. So FX is based in a simplistic manner for our audience to understand. It is a cost to convert currency A to currency B, right, at any given point in time. The spread between these two currencies is what we call FX, the foreign exchange rate. Now, the problem with this industry is the FX spreads for any currency is intentionally designed to be very opaque.

4:36when you actually try to send money from your bank account. You say, okay, you get this rate and you send this currency, you receive this rate. The spreads are not shown, right? Even if you actually go to transfer-wise and if you try to basically convert currency to B, they say this is the best rate. But to understand the spread between the bid and the ask, you'd have to try to reverse the order and see to sell the same currency and buy it, what exactly does it cost? Then you actually get the bid-ask rate and realize, oh, for a$100 ,000 transfer, I'm actually paying about 2 % to 4%. right even on some of the best platforms in the world that is a bit as great that they make now does it actually cost that much no it does not what happens is each of the infrastructure providers use various underlying service providers and each service provider tax on their own margin and spread why because there are structural challenges it's not like everyone's a greedy bastard who's trying to make a lot of money it's just that there is a lot of infrastructure costs along the way.

5:30Now what has happened in 2025 is a few things. Number one, stablecoins have found PMF in the last few years. What that has enabled us to do is bypass infrastructure that we were not able to bypass before, namely SWIFT, right? Because SWIFT is how you send money across borders historically, but you're able to now bypass SWIFT completely via stablecoins, significantly reducing the cost structure of moving crossborder.

5:55Prabhakar Reddy: So there are really two components, it sounds like. One is actually the conversion of one currency to another. And then one is effectively a payment. And often these two, as in a transfer of currency from party A to party B, and often those things are happening simultaneously. Is that right? That's correct. And how do you think about those two different components of the flow? Is one inherently harder or easier? Yeah, very good question. And actually, there are more legs to a cross-border payment than just converting and sending. If you look at it, there are about four different legs. First is collection of a payment.

6:33When you're selling money to another person, there's often an intermediary fintech or a bank which needs to receive that money. So collections in itself is a very large business. There are multi-billion dollar companies, publicly listed companies that are purely in the collections business. Because when you have to integrate with various different wallets, various different currencies and various different geographies and simplifying it, It is a complex business. So collection is business number one.

6:56Prabhakar Reddy: What are some examples of those businesses? I'll give you an example of a publicly listed company in the UK, actually. It's called Boku. They've been running around for 15 odd years. Phenomenal company. Very good team. Some of the largest FANG companies. Everyone used them for collecting money around the world. They charge, it's all public information, so you can look it up. They charge roughly about 80 basis points for collecting money in various different currencies. Just collections. and once money is collected in currency A, it needs to go into a bank. So there is a back in the bank. Just on collections more, a little deeper.

7:28Prabhakar Reddy: You think of, you know, Stripe for online payments or some sort of like card, you know, processor who's basically allowing someone to pay with a card and delivering those funds to a merchant's bank account or something. This is like a more EFT, like electronic payments version of that, where they're facilitating maybe like an ACH payment if they're a U.S. company or they're facilitating like a faster payment if they're in the U.K.? So when you look at G3 currencies, world looks differently. But when you look at a lot of the emerging nations and the number of wallet integrations, the number of banking type integrations that you need to do, things get more complex, right?

8:08So if you want to collect money in Mexico, you cannot just do a bank integration, you have to do a SPA integration. You'll have to do some wallet integrations. You want to collect money in rupee, you'll have to basically do an UPI integration. You would have to do a bunch of like phone pay, Gcash, lots and lots of integrations in every country and currency. And now suddenly you have a very bloated infrastructure stack. If you want every company is trying to replicate it, which is why companies like Buku exist, which simplify the entire experience and ensure that you have 99.999 percentile success rate and collection of money.

8:37Prabhakar Reddy: Yeah, makes sense. OK, thank you. That was collections. Yeah, that's collections. Once money is collected, it needs to go into a bank account. Now, there are two ways to collect money into a Swift international bank transfer account or a local currency account, which is within the country itself. Creating that infrastructure is very, very complex. There are companies like CurrencyCloud. There are companies like Airwallex that have popped up saying, we'll give you infrastructure to just manage banking infrastructure in a simplistic manner. Those are wallets, but truly not last mile accounts. There is no global last mile infrastructure for every bank account in every currency.

9:07the closest you'll probably get to is JP Morgan which probably has like 80 currencies and 80 countries covered but that's about it if you really want to open you know a Mozambique dollar bank account you'll have to go create a company in that country get a local bank account local infrastructure local director and it takes six to nine months per country per year so that is the amount of time it takes to just set up the banking infrastructure now so you've collected a currency A it's gone to a bank account A now you need to have the FX select so you need to convert the currency from currency A to currency B.

9:37So that FX infrastructure, again, it's a very complex piece. Now that's the next step. Once that is converted, it needs to get converted to currency B. You need a bank account for currency B as well. So we're back to banking infrastructure, right? Because you collected pounds, you need to convert into Filipino peso. If you keep it in Barclays in your Filipino peso account, for you to send it into your local Philippines infrastructure will again require Swiftwire. So preferably you need to have local Philippines bank account, right as an example so that's the banking infrastructure once that's done you need to have the last mile payout infrastructure as well if you need to pay out again in philippines via gcash you need to have a gcash payout integration if you need to send via swift transfer somewhere you need to have swift integration if you have any other local payout infrastructure you need that integration so every country requires its multitude of payouts infrastructure so again there are multitude of companies that have been built and just payouts mastercard move visa direct all All of these are payout infrastructure.

10:30So every single leg of this is a very complex thing. Now, any fintech, if you're starting a fintech tomorrow, you're a YC company, you will have to basically integrate with a multitude of collection infrastructure, a multitude of banking infrastructure, lots and lots of FX infrastructure, lots and lots of payout infrastructure just to do global cross-border development. That's how complex this world is today.

10:51Prabhakar Reddy: Wow. So I can see then why, you know, startups might be excited about the idea of, well, hey, hey, let's just use some wallets on the back end. Everything's stable coins. And now it's like international Venmo, peer-to-peer payments, like problem solved. Tell us why, you know, about the disconnect kind of from the current, not only market structure, but the way that, you know, businesses use payments, use these counterparties and, you know, that sort of utopian vision of like, everything's just a P2P stable coin. So stable coins have come out in the last few years for solving a very different problem.

11:27Businesses and SMBs don't use stablecoins natively today because their last mile recipients are not yet ready to accept stablecoins as a form of payment. In very exceptional case, again, maybe one to 2 % of the cases, you do have businesses that do, but 99 % of the world, your coffee shop around the corner is not going to accept stablecoins for payment. They really need dollars or euros or pounds or whatever currency you're living in, right? So you need to basically deliver that. And so even if your fintech app is going to allow you to pay by stablecoins, that merchant who is at the end needs to basically receive that fiat currency.

11:59So there needs to be a stable account of fiat conversion and delivery of fiat to the end merchant, which again requires in the fintech in between, which is again a complex infrastructure because depending on the fintech, they can take two, three, four days or however long to basically send that. Again, very complex infrastructure. All of this is what we are trying to grossly simplify at OpenFX. What we're trying to do is create one API that collects money, banks money, converts money and pays out money in an extremely seamless manner for any company, whether you're a fintech, whether you're an e-commerce company, whether you're a neobank.

12:33If you want to create the next revenue tomorrow, you don't have to go through 15 years of infrastructure building. Literally get up and running by using OpenFX infrastructure for all of this. That's what we're trying to sell. Build the underlying AWS rails for global cross-border money.

12:46Prabhakar Reddy: Do you think anyone else has ever, and I know many other startups maybe that are contemporaneous to you, but, you know, for the past 15 years of fintech, has anyone else ever pitched that vision? And if so, like, is it different this time? Or why didn't it work last time? Or what are you saying that's different from what they were saying in the past? Fantastic question. And I've been a VC in the past. I've worked with Excel. So I can apply both lens. So let's go back to Generation 1. Generation 1 are Barclays, JP Morgan, HSBC, Standard Chartered, Citibank. They're what you know as transaction banks.

13:21So globally, most of the cross-border money movement happens through about the seven to nine banks around the world. They process most of it. They do have the infrastructure, but at the end of the day, they're built on infrastructure that was built in the 1970s. So lots of Nostro-Wostro relationships. Again, misaligned incentives.

13:37Prabhakar Reddy: Banks want to hold money for as long as possible because they make interest margin on keeping the money with them. Can they pay out money in every country? Technically, yes. But what's the incentive? Because if you look at the balance sheet and you look at the revenue structure, you realize that most of the money comes from keeping the money with them as long as possible. So that's generation one. Then came generation two, starting in around 2011, starting with a company called CurrencyCloud, which pioneered this entire category called Embedded Finance. And then came Neum, Airwallex, Thoons, Dlocal, and a lot of companies across different parts of the world.

14:09What did they try to do? They said, okay, for the next generation of fintechs that are coming in and all the technology companies that are coming in, it's impossible for them to go and integrate with the JP Morgan and HSBC and Citibank and every single bank around the world. It's complex. We're going to API-fy all of this banking infrastructure and give them an API, right? So that you can actually collect money, bank money, convert money and send out money through them. Wonderful. It helped the world get to where it is today, right? But the problem with the infrastructure is when you're built as a wrapper around Barclays, Barclays does not move or work.

14:39after you know bank cutoff times which is 4 or 5 p.m in the country barclays doesn't work on a weekend barclays doesn't work on a national holiday barclays also has to send money to let's say bank of philippines in philippines and they're going to be two to three nostril-ostro intermediary banking infrastructure relationships between them so now bank k has to send to bank b bank b whenever they wake up in their country in their geography in their time zone we look at it then process it and against all humans and operations and all of that stuff it takes a lot of time for money to finally reach the other destination.

15:07Prabhakar Reddy: Now, if you basically look at where the world is headed to it, at the current moment, we're in a very interesting inflection where on one side, you have fintechs that are kind of used to near instant payments because of stablecoins, expectation is increasing. We're also entering this world of AI where agentic payments and AI agents are basically going to expect to be doing a million transactions of one to two dollars each at any even point and want to basically receive the money on the other side near instantly. They're not going to wait three days or five days to basically receive money to process their transaction.

15:42Expectations are changing. The world order is changing and everything needs to be programmatic. Previous generation, not programmatic. A lot of the banks, even today at OpenFX, we actually have to create RP, robotic process automation systems to actually just create an API on some of the traditional banks. So that is where the previous generation is. So the world order is changing. The expectations are changing. And the current expectation is that you have to be able to process a$500 million transaction in clip sizes if required of$50 each and settle that in under a few minutes so that the other party on the other side, even if it's Saturday at 3 a.m., is able to receive it and get the job done.

16:18So that is infrastructure that we are building towards.

16:22Prabhakar Reddy: Wow. All right. So you talk about kind of like rebuilding the whole stack or building the whole stack. Talk more about that. But I guess there's one world in which you build a completely parallel setup and you're doing everything yourselves. And there's probably, you talked about what the currency cloud world was, where you're kind of wrapping the existing infrastructure and delivering an API on top. And then there's probably something in the middle. Talk to us about what you mean by building the whole stack. There are several layers of the stack that function efficiently. For example, do I really need to build another Stripe?

16:53I think Stripe has done a fantastic job, right? And every Canadian has done a fantastic job in Europe. and Razorpay has done a fantastic job in India and so on and so forth. Those layers of the stack, I don't really read it. But integrating all of that into a seamless API and offering that with 99.999 percentile efficiency for all the payment companies and fintechs to get access to collections is an opportunity, right? Doing that, that's one. Number two is banking. Do you have one API that actually can offer you global virtual accounts and banking in a seamless manner without charging an arm and a leg and a$50 million deposit to get access to all this infrastructure with almost an internal CLS kind of system where you can settle money 24-7?

17:33There isn't. There is a great opportunity to rebuild the entire backing stack. FX, to literally just get access to African currencies, for example. There is no APIs for most of the African countries. You have to go onto Telegram or WhatsApp and basically request for quotes. How do you programatize all of this stuff? How do you ensure that you're offering a price guarantee for an e-commerce company that needs to sell their goods and list it without having to change their prices every two hours because of fluctuating rates.

18:00Prabhakar Reddy: There's a lot of infrastructure that needs to be literally thought from ground up. That is what I'm talking about, right? Realize that across this entire, you know, the four legs that I talked about, which is collection, banking, FX, and payouts, the most complex piece is FX. Collections, independently in every country, there's good jobs being done. Payouts, there's a lot of good job being done independently in most geographies. Banking is a very complex piece to solve, but FX is the worst of it all. The most opaque industry, there is no transparency. prices are all over the place there is no consistency across so that's where we decided to basically plant our flowers start basically building infrastructure from ground up own that entire piece and then start building adjacent products right because banking if i can pipe now with the barclays or jp morgan and offer that infrastructure to the next fintech that's starting in a wrapper that's fantastic because for them to access jp morgan the banking infrastructure will take five years before they even get onboarded onto the platform as a yc startup but if i can offer that in two clicks of a button at zero cost, that's fantastic cost infrastructure.

18:57Prabhakar Reddy: Yeah. So the core, the money conversion piece, that's where your Falcon X heritage comes in, right? I understand that you're effectively building an FX market making function at the core of OpenFX. Can you talk more about that? Absolutely. So look, there's two ways you can build an FX desk. One, you can wrap up, wrap your infrastructure around all the existing banks and OTCs and market makers. But you only hit so much scale at the end of the day because you're bound by their gravitational force, which is their balance sheet, their settlement times, their operational infrastructure. So if you really want to build a company that can process a trillion dollars a day off FX moment, realize wrapping your own infrastructure doesn't work.

19:39We have to build infrastructure ground up. So an analogy is we are Robinhood, but we are also Citadel Securities underneath that actually provides the infrastructure for market making. So in every currency, in every corridor, we actually set up our own internal market making infrastructure so actually we hold positions we understand what needs to be done because unlike crypto which you know was an industry that was previously in which moves you know maybe all coins move about 20 every two minutes whereas fx moves maybe two percent in a month four percent in a month at the very most so it's a lot more controlled risk that we can actually manage so that's one number two which is very interesting over here when you're building market making infrastructure for fx is we have continuous flow across lots of fintech companies So we actually have built a predictive ledger.

20:22We know to 85th percentile accuracy at any given point. And today we're recording this podcast at 9 a.m. on a Wednesday, PST. I actually know at 9.30 a.m. and 10 a.m. and 10.30 a.m. and 11 a.m. across the world with 85th percentile accuracy. What kind of flows can I expect historically and actually predict and place my balances accordingly so that I can deliver currencies to my customers in 60 minutes whenever required. That allows me to lower my cost, net out my margin, because I also know that preceding hours, what kind of flows can I historically expect? I can change my prices accordingly. That really helps me bring my cost structure down, manage my infrastructure about better.

20:59Prabhakar Reddy: Yeah. How much capital is required to run a trading desk to process a trillion of notional per day? You know, in a very inefficient world, you'll need a trillion dollars of capital. But that is not scalable. So it really depends on how much you can rotate your capital at any given point, right? Because what you have to look at, your peaks in any given hour, right? So if you're processing, today OpenFX processes roughly$50 billion a year, which comes around to about$4 billion a month, right? $4 billion a month means we have peaks of about, on any given day, maybe$500 million at times. For that, a balance sheet of about$150 to$200 billion really helps scale it, right?

21:42Again, there's a lot of efficiency that we can still further derive and we can bring that to a 5x rotation piece where$100 billion can get us to$500. We are currently, you know, at that scale.

21:52Prabhakar Reddy: Yeah, well, I imagine a huge part of it is ultimately netting and settlement times. And presumably the simple fact that you're looking to guarantee payouts within 60 minutes and, you know, who knows, maybe two, three, five years from now, payouts within one minute has meaningful implications on the capital requirements. Can you unpack that a little bit more for us? It does. So the thing is, there is economies of scale that really start coming into effect because economies of scale help you basically understand your flows a lot better. And at scale, you can also sort of get the right kind of customer persona that balances out your flows.

22:31So, yes, a lot of internalization happens for OpenFX today because we incentivize our go-to-market team to say, okay, we're seeing a lot of, you know, one-sided flow on, let's say, Mexican peso. So for the next quarter, please go bring the counter flows, right? For example, when you have a lot of remittance companies for Mexico, they're sending Mexican peso into Mexico. But if you start getting, let's say, trading companies that are actually purchasing equipment in the US, they are sending a Mexican peso out. So you can actually balance those flows out. So trying to mix the personas of clients for every region is also a complex thing.

23:04So you have to combine your market-making infrastructure with your go-to-market strategy, with your product pricing strategy. because in real time, you can actually also ax your prices, so be in your holding positions, and you get the right kind of customer persona, which is like, say, market makers and sharks. You will actually take advantage of the small price axes, even if it's a 0.1 delta, right? It's advantages for them. For us, it's very profitable. Rather than going and hedging off on like a Bank of Mexico or Sandander, we can actually just hand it off to a market maker

23:29Prabhakar Reddy: who will take advantage of that small skew and pricing. So combining all of this in real time, programmatically, second by second, is very complex. And that's what. Yeah. Yeah. So I imagine you're this market making function. It's not exclusively processing transactions for your own customers. Like in many cases, potentially the other leg of the transaction is against some traditional FX desk or market maker. Is that right or no? So, you know, when we talk about the volumes that we process, all of that is customer facing volume. Everything else that we have internalized is often 10 times more than that.

24:06like everything else that we do to hedge off with our liquidity partners, market makers. There's a lot of behind the scenes volume that gets processed. But again, that's not what we talk about or count as well, because all of that is, you know, how the sauce gets made. But at the end of the day, what really matters is how much value we're delivering to customers, which is a non-star metric, which is a TP meet, which is customer facing.

24:25Prabhakar Reddy: So, I mean, it sounds like in a sense you're building two businesses. One is this FX market making business, which presumably could be a standalone business of its own. And I mean, you talked about basically like the importance of automation. So I assume it's nearly fully automated, which must make it stand out, presumably compared to at least some of the competition. And then also this sort of global, you know, money collection, transfer and kind of last mile delivery network. Are those the two parts of your business? Is there is there more as you think about it? At the moment, those are the two parts of our business because we started with FX as a core anchor.

25:02Now, I want to create a call out an important distinction here, whether it's Barclays or J.P. Morgan or Standard Chartered. Everyone runs very similar infrastructure, but I think it's very important. And this is very important to me. We have a very clear mission vision for why we started this company and why we're building this company, because it's very easy to get lost in this. Oh, we can make a lot of money using my market making for effects. So let's actually, you know, squeeze more money out of here. But that's not what we're here to do. We are actually here to collapse this entire FX industry, bring it down to such low margin infrastructure that we can actually run a global payments business while running a reasonably profitable FX infrastructure business.

25:39So that's a small twist here, because if we lose sight of that, it becomes very easy for us to say, hey, we can actually charge 50 basis points for this Nigerian currency. Why don't we actually start doing that as opposed to we can actually profitably run this at five? Why don't we actually just charge five, six or seven basis points to our customers and make the world a better place? So that's the small important distinction that I'm going to call out.

26:00Prabhakar Reddy: Well, I've heard you talk about, you know, kind of coming in with new currency pairs and almost like collapsing market pricing by an order of magnitude. Is that true? And is there anything you'd like to share about that? Yeah. In every geography that we've entered, we've actually brought down FX pricing down by 80 to 90 percent to the point where now when OpenFX announces that we're entering a new country, the existing players, and this is actually happening in every geography, they actually scramble and slash their prices by 50 to 60 % just to remain competitive against open effects. This shows how much margin there is left on the table.

26:37But even then, we enter those countries, still profitable in those countries, and we are able to just strip the market out because there's a lot of money being left on the table. Give you an example.

26:48Prabhakar Reddy: AED is a good example in UAE. So UAE therms is a currency of UAE. before open effects entered their geography the spreads in the market were roughly 40 to 50 basis points and this is again a pegged currency it's pegged one is to 3.675 by the government to the US dollar these are the spreads we were like this is ridiculous this is a lot of money just being left on the table cut the story two years later the spreads are now single digit basis points less than five we've literally brought the entire market down and you go into the market you ask hey where can we basically get the best 80 spreads, ability to open effects, hand it down.

27:22So that's the impact we're having in every job.

27:25Prabhakar Reddy: Amazing. All right, so you have experience as an operator at Falcon X. You have experience as a venture capitalist at Excel. What has this been like? I mean, you know, you had spent a lot of time, like early stage, you spent a lot of time building. Tell us about the experience. What's gone well? What's been hard? What surprised you? Where's your team even based? How do you guys operate day to day? very good question um so open effects is my fifth journey as a founder i've been a founder for about 20 years now it's all i've known for a very long time so the buildings things from zero to one every single journey has been very different the fifth time around the advantage you have is you've made a lot of mistakes in the past four journeys and you can course correct them uh i make i make mistakes all day long but they're new mistakes not the same mistakes We say the same thing.

28:15We say the same thing. This journey has been a lot more fun for me this time around because I'm building the company very differently. So one of the first things I did very intentionally before I even incorporated the company, I was actually put together a culture doc. Because I realized where I want to get to is very important to me. But how I get there is equally more important. Because I don't want to wake up four, five, six years from now and be like, what is this? What is this company that I've created? I can't breathe anymore. right that's i don't want to create that kind of environment not for me for or from exec team or for anyone else so culture is very important that's been one key thing number two that i'm doing differently this time around is creating a very robust exec team bench from day one because what founders usually do first-time founders do is create a senior exec bench much later in their journey because they're not confident whether the business will get there whether they need these kind of people but having been through the journey i know the advantage of bringing folks who've been And they've done that for various different functions.

29:13And I'm very proud of the exact thing that we have. That's how we've been able to get to the scale that we are at in about 24 months of launching. Again, comparatively, WISE, TransferWise, which you brought up earlier, processes about$250 billion a year with 6 ,500 employees. We process$50 billion with 25 % of WISE processes with 100 employees. And most of these are engineers and quants. And there's a reason we're able to do it at this scale and grow not in here. Three, you asked about where is the team based. I build my team globally around my leaders. So we actually split in four different geographies, Miami, London, Dubai, Bangalore.

29:48So these are the four hubs that we have. And this is where my leadership team is based. And I keep shuffling between all four.

29:53Prabhakar Reddy: And everything has just been, you know, a straight line from initial vision through execution to where you are today. You know, when you zoom out far enough, everything looks like a straight line. But as you know, you're an operator yourself. It's never a straight line, right? There's lots of ups and downs, But in the grand scheme of things, those are downs that are always required, right? Those downs actually help you pull you back so that you can leapfrog much faster. What have been the downs? Let me talk about a few. You know, I've been through this licensing regulatory process in the past in various geographies, but the complexity of having global licensing footprint is not as simple as you can imagine, right?

30:33Like every regulator has the nuances. So it is a lot more complex. The timelines that you think you're going to accomplish things in often take two to three times longer because of unknown factors, right? What can you do when you're in the UAE and you're getting bombed in a war that you didn't plan for and nobody's working, right? Do you think slow down naturally, as an example?

30:53Prabhakar Reddy: That's one. Number two, I would say banks. You know, I honestly thought the biggest challenge is going to be onboarding and getting access to the banking infrastructure. I didn't even comprehend how behind a lot of the international banks are, where most of them don't have APIs. A lot of US banks don't have APIs, right? They refuse to give APIs as well, a lot of them. They're like, oh, we'll see, maybe in nine months, 12 months' time, we'll open up our APIs. The API documentation is crap for most of the time, even now and they open it up. So we have to create, as I was mentioning earlier, RPA systems where we have, because we can't process things in minutes if we have to throw humans at the problem.

31:28To automate everything, we have to literally mimic our humans and create our own APIs on every bank infrastructure. That is like a thing that caught me by shock as to how behind the global banking infrastructure is. The third thing which, you know, I would say caught me by surprise is how long it's going to take to basically build infrastructure that is going to process a trillion dollars a day. Because FX moves seven and a half trillion dollars a day, effectively, right? I believe that if you're able to bring down the cost structure and the timelines of moving money from days down to single digit minutes, the entire pipeline of cross-border money moment increases.

32:06We were naive and when we started, we thought we'll get there in about five years. And I think if you get to a trillion dollars a year in about two to three years, that's a big milestone. Getting there from a trillion dollars a year to a trillion dollars a day is, again, a decade's journey, in my view, right? Because a lot of battle needs to be fought. So the timeline, the complexity has also been quite a surprise for me.

32:26Prabhakar Reddy: Well, maybe to wrap up, you can look forward a little bit. So picking up from there, you guys are obviously building an incredible business. What do you see happening around you in the market? And I don't mean direct competitors, but on one side, you have potentially what seems like increased diffusion, distribution of stable coins, potentially blockchains as financial infrastructure. On the other side, you know, it's easy to look back at many different hype cycles around new banking or financial infrastructure and realize that actually incumbents operate, you know, basically the same way that they always have, perhaps with slightly compressed margins.

33:03Prabhakar Reddy: But nothing, you know, fundamentally has really changed. How do you think about what this market looks like five years from now or 10 years from now? I think five years is a difficult time frame to highlight because a lot can happen within five. But in 10, I can say with high degree of confidence that the world is going to look very, very, very different from what it does today. Specifically, since we're talking about cross-border payments, I'm going to limit my comments and statement to that. Everything in cross-border payments that was built in the last 50 years has been built around human timescales.

33:35It's built for end-of-day batch processing to be done by human beings. In the next decade, that's going to completely change. I think AI and AI agents are going to completely change the way we expect money to move, the way we collect money back, money convert or pay out money. It's going to completely change. The expectations around scale, time, cost, efficiency is going to be radically different. Today, the world is built around processing two transactions at the end of the day of a million dollars each. In a decade, we're going to process or the expectations are going to be that you process a million transactions of two dollars each per second.

34:10Prabhakar Reddy: Right. So that is the scale at which things are going to operate. The world is going to look very different. Quite excited, to be very honest, for that kind of world. We're building towards it. The world is not ready for what's to come because most of the world's banking infrastructure, corporate infrastructure is not ready for it. So I think the largest companies in a decade are going to be very different than the companies that we see today. Amazing. Prabhakar, congratulations again on everything you've built. Very excited about watching your progress from here. Thank you so much. Thank you for having me, Abhil.

34:40Thank you.

35:10Thank you.