Why Ramp Is All in on Stablecoins

Alex Bazhenov  0:00  

The way that I think about our stablecoin product, or the way that I think about the future of it, is I hope that in a year or a year and a half, two years, we're not even talking about a ramp stablecoin product. Our goal is basically to make it easy to hold money on ramp and make it easy to use money on ramp. And I think that right now we're at this kind of like inciting inflection point where we can talk about stables on ramp. We can talk about stablecoin accounts, but really the goal is to abstract all of that away. Welcome to Tokenized. The show focuses on stablecoins and the institutional adoption of tokenized real-world assets. My name's Simon Taylor. I'm your host, author at FinTech Brain Food and head of Market Dev. Tempo and joining me as always is Rand Goldstein, Goldie, SVP of Payments and Network at Fireblocks, Goldie, my friend. How the heck are you?

Ran Goldi  0:47  

I am amazing. I just so waiting for this story about Ram. There's so many questions we're gonna ask them, and we we did send them a lot, but I will already say that we've changed all of them. So good luck to all of

Sy Taylor  1:00  

us. That's what you want. There are very few guests who come to a conversation and go, you know, like what if we did this instead? And that's my favorite kind of guest. And my goodness, do we have some great guests for stable kind stories today? Just before we jump into it, why are we doing story time again, Goldie? Just give me the format of the show.

Ran Goldi  1:17  

Well, it's a good question. It's it's really because if you go around the world today, you'll see that it's filled with news about stablecoins, with PRs about stablecoins. But we know that that's not the real story. People want to know what's happening behind the scenes, what's happening behind the companies that are redefining these tectonic shifts, and people want to know what we're going to tell them today, which is the story behind the company that well, you could say they were just doing spend management or card issuance, but they actually redefined how AI and stablecoins come together. Today, we're with Ramp.

Sy Taylor  1:51  

Indeed. So, Alex Bazanoff is senior staff software engineer at Ramp. How you doing, Alex?

Alex Bazhenov  1:57  

I'm doing great. How are you doing,

Sy Taylor  1:58  

Simon? I'm really, really well. Been enjoying your tweets or ex eats, whatever they are these days. Excited to have you on the show. A good friend, Andrew Chapello, product lead at Ramp. How you doing, Andrew?

Andrew Chapello  2:10  

Hey, doing great. Really excited to be here.

Sy Taylor  2:13  

Good to have you on the show finally. Just before we get into the content, I got to remind viewers and listeners that views and opinions of our contributors today are their own and might not reflect those of companies they represent. And please don't take anything we say as tax, legal, or financial advice. This is for educational purposes only. And a reminder, of course, that this podcast is made possible by Fireblocks and our good friends at Visa. All right, so we're going to talk a lot about REMP today, but before we get into the weeds of that, Alex, can you tell us a little bit about you and your journey to getting to staff engineer at Ramp doing stablecoin things?

Alex Bazhenov  2:50  

Sure. Yeah. So yeah, I am an engineer at Ramp. I head up all of our digital assets engineering efforts. My journey started in 2018. I was a early engineer at ZeroHash, so very much living in that kind of niche of tradfi and DeFi from the very beginning. Worked at a number of different companies between then and now, between kind of like high-frequency trading firms, venture studios specializing in DeFi, and then most recently at Bastion, which is a kind of issuance and stablecoin as a service company, I think listeners to the show are probably familiar with it. And yeah, joined Ramp about just under a year ago to kind of really build out the stablecoin and digital asset effort here, zero to one.

Sy Taylor  3:35  

Andrew, give me your backstory, and we'll get into the the Ramp side of it shortly on the stablecoins. But how do how do we find you here today?

Andrew Chapello  3:43  

Yeah, I'm a long time fintech nerd. Have built a billing system at a company called Recurly. Built a one click checkout platform that was called GoCart, and ultimately came to Ramp to help sort of lead their their payments efforts. And so I'm newer to the digital asset space, but came into it in in that we saw stablecoins and digital assets as a way to solve very clear and present problem for our customers. So excited to get more into it today.

Ran Goldi  4:13  

I think it's actually interesting, Andrew. That again, as you said, you you know obviously modest, but you've been to several super interesting businesses before, right? Like Box and Recurly, and you did not come from this space. Like you know, Alex has been immersed through ZeroHash and other companies. Simon has been for many years. We've sort of got accustomed to this thing called stablecoin. We don't even question it, right? But when did really Ramp got started to get serious into stablecoins and into tokenization and to actually that adding value to your business? And I think more importantly, what I want to hear the story is like, did you really believe in this thing, or did you come and say, well, I don't think we should touch these assets, but eventually actually found out it's amazing or not?

Andrew Chapello  5:00  

Yeah, it's it's a great question. I think it like everything around starts with what's a customer problem that we want to solve. As we we were spending more time talking with customers, especially those that were operating around the world over the last I don't know two years or so. I think early last year we started to come up with a hypothesis that the the types of solutions that we had were not good enough, and the way that we got there was we were working with businesses like Airbnb and like Shopify and Uber that were operating all around the world, and we had the same complaints coming up over and over again, why does it take five days for me to move my money to the to my vendor in Europe? Why does it cost you know maybe one, two, 3% of the transaction cost to be able to move move my money this way? Why are there so many different delays in the banking calendar? Whether it's bank holidays, cutoff times, it was it was clear that the ambition that we had in in moving money, whether it was for sort of corporate treasury use cases or for paying employees or paying vendors, could not be met with the sort of traditional banking infrastructure, and as as we started to peel the layers back to the onion, I sort of kept coming back to this problem of you know people talk about the global financial system like it's something that was premeditated, something that was designed, something that you know is is like sort of centrally planned. Oh, sweet,

Sy Taylor  6:39  

summer child,

Andrew Chapello  6:41  

yeah, yeah. It's it's and like you all know that that that doesn't exist. We have 150 sort of poorly linked together financial systems, and and and you know the the our entry into digital assets really stemmed from this key insight of like, what if there was a better way to do this? What if you were to design this from first principles, a system that was truly global by default, that let you move money at will at any time, that that that reduce cost, that really lowered borders, and and so you know we started looking into this in detail. That's probably last year. That was our inception.

Sy Taylor  7:20  

And so, take me under the hood of that, Alex. The internal Ramp story is this like you and Andrew going around stablecoin pulling people? Is it more of a customer-led thing? What does that look like, and where do you come into the picture?

Alex Bazhenov  7:34  

Yeah, so I think that on some level, everything that we do at Ramp is customer-led. We're a business that's like, I mean, you probably hear it all the time entirely about saving time and money for our customers. So I don't think we do anything that is not like fundamentally aligned with what they want. My experience here has very much been that the entire organization, top to bottom, is aligned on stablecoins, aligned on the vision, aligned on the future of money movement and money storage. I know that there are a number of institutions that are that are getting into stablecoins and have been getting into stablecoins over the past like year, two years, especially since the Genius Act passed. And there are definitely battles that are being fought within those institutions to you know stablecoin pill or crypto pill or whatever you call you want to call it, that's very much not the case here. So I I came in after Andrew, and my experience has not at all been like okay, like I am the stablecoin guy or I'm the crypto guy. Let's let's you know like convince these people that there's market need here. Let's convince that there's our customers or internally that there's some kind of like product market fit here. I think it's very clear internally and externally from feedback from our customers that like stablecoins are the future of Ramp.

Ran Goldi  8:50  

So I so let's let's double click into that. First of all, Ramp. You know I think Ramp is a lot of things. I know that today Ramp goes by the title of you know finance AI platform, sorry if that's the right way to say it. And again, you're not public yet. I don't know why you put AI in the name. You could have wait until you're public, then you would get like you know 100 push on the stock, right? But that's what Simon and I are doing with with tokenized stories. We're just waiting for us to go public on this, but you've you've grown like the company has grown more than 156% growth. By the way, in terms of like people headcount over the last two years, which is phenomenal, especially for a company that again this is not like 100 people. This there's more than 2500 people right now. So if you can take us through, I guess first of all, like what are the top things that that Ramp is doing altogether? But then if you could, Alex and Andrew, if you can touch like specific two or three stablecoin related products and what, as as you said, Andrew, what problems are they solving today? And I want us to then tell the story behind that.

Sy Taylor  9:58  

Yeah, and I think where it fits. The hierarchy of the overall like ramp product set because you do a lot of things. People see think spend management card, but this thing is a whole lot more than that.

Andrew Chapello  10:09  

Yeah, at the highest level, our mission is to save businesses time and money, and and we are interested in doing anything that helps them solve that. And so, you know, we've got a few sort of high-level products, and we're sort of delving into a few new areas that I can tell you a little bit about and how save points relate to those. So our our original and and sort of still our core product is our spend management product. That is, you know, the corporate card that we're very well known for. That you swipe it at a merchant and immediately get a text message to categorize what was this for. Send us a picture of your receipt, and then your expenses are done. You know, seconds after the transaction happens, we still are investing really heavily in that product. That's still probably what we're best known for. We also have a really, really strong procure to pay product. So that being both a procurement of I want to I want to buy something for my company and sort of running that through internal approvals, getting your budget approved, selecting vendors, responding to RFPs, those kinds of things, and then the other side, which is paying your vendors that you've now selected in bill payments, and that is largely bank to bank payments. It involves some credit card payments as well. I didn't mention within our our sort of expense management product, we also support reimbursements. So if you are spending money on behalf of your of your company on your personal card or through your personal bank, you can get repaid through your blueprint. Beyond that, we have a travel product, and we have a lot of like really interesting emerging bets around accounting. We released. a product called Ranch Stack several weeks ago that is is focused on trying to as much as possible automate your month end close, and so you can give Stack your month. Does that

Ran Goldi  12:15  

include your crypto related wallets, your yield, your staking wallets, your Morpha Vaults, or is that still in the fiat world?

Andrew Chapello  12:24  

I think it's mostly still in the fiat world, but in theory, it can include anything. You know, it's it's it's sort of an AI generated product to be able to automate your process. And so, like, if if your team at Fireblocks had a month-end close process that you know is probably 1000s of lines in a spreadsheet, you could give that to Stack, and Stack would automate that for you, and like help to make it make each step of your process, whether it's accruals, whether it's you know being able to to do your your like amortization schedules, we can we can handle all that.

Ran Goldi  12:56  

So Andrew, would it be would it be right to say sorry, hitting the chart, but would it be right to say that for every financial product you're doing at Ramp for the fiat stack, you would also eventually do that for the stablecoin stack, or are you looking at those as two different worlds?

Andrew Chapello  13:13  

I think it is fair to say that everything that has a fiat equivalent has now and and will continue to have a stablecoin equivalent, and we think that by offering a stablecoin and by offering digital asset products, that we can offer that to more customers and we can make it more useful. So, for example, within our our bill pay product, you know you're paying vendors around the world. Say you have a vendor in say sub-Saharan Africa or in Argentina, and they don't necessarily want to accept their local fiat currency, or they prefer to accept stablecoins. You can have a better relationship with them if you're paying directly in digital assets, and you shouldn't have to care about like needing to hold stablecoins yourselves. We allow you to do that, but our goal is to make it really easy for you to accomplish the job that you want to do, and if that's paying the your vendor in their preferred currency, that's great. If it's you know I have a a digital wallet as a business and I want to move funds from my entity in the U.S. to my entity in Dubai, that should be easy. That should be seamless, and we want to like add that on to all of our our existing products. And the one that I didn't mention that that's this is really tightly integrated with is our banking product. And like up until earlier this year, that was fiat focused product. You could hold USD, and we could help you earn on your balance as well as meet your obligations at the time when you need it to, and we see that as as interchangeable with stablecoins, and that's a lot of what we've been building.

Sy Taylor  14:48  

Yeah, that sort of stablecoin account, that CFO dashboard, where it all looks the same to me. Whether I've got stablecoins in there, whether I'm paying a vendor, whether it's fiat, I can just do it whatever way I need to, and I've got this single. Pane of glass to kind of kind of manage all that. So, Alex, I don't know if you know this, but can you talk me what take ups been like and what it was like to build this stuff under the hood as well?

Alex Bazhenov  15:10  

Yeah, I mean, candidly, we have gotten like immensely positive feedback on our stablecoin product. A lot of that has to do with the fact that I think a lot of businesses don't really know the options or have even have access to the right tools to conduct their finances, their AP, their AR through stablecoins. Of of the companies that have onboarded, we have companies that are onboarding from using a hot wallet that's just like a MetaMask wallet, effectively to run the AP of their business. Which I think no shade to MetaMask, but from like a like an organizational perspective, right? I think that that's like a little bit of a of a nightmare. We have businesses that have kind of like clobbered together their own solutions. So I think that right now the feedback and the uptake has been like very positive. But candidly, I don't think that there's like very strong alternatives in the market. The way that I think about our stablecoin product, or the way that I think about the future of it, is I hope that in a year or a year and a half, two years, we're not even talking about a ramp stablecoin product. Our goal is basically to make it easy to hold money on ramp and make it easy to use money on ramp, and I think that right now we're at this kind of like exciting, exciting inflection point where we can talk about stables on ramp, we can talk about stablecoin accounts, but really the goal is to abstract all of that away. What we're building towards is just dollars or euros or whatever kind of like asset or whatever value at rest in ramp, and then building the interconnectivity to stables, two different pay-in rails, two different payout rails. But yeah, that that was that's like a pretty roundabout way to just say that uptake has been great, the response has been great. We are very much looking at both feedback from kind of like crypto native companies that are onboarding and the companies that are just Main Street every day. Well, we have like farmers that use our stablecoin product as an example, and across the board, it's just been like, yeah, quite positive. And honestly, like not as difficult to build with the like level of partners and tooling and maturity of the like the product offerings and engineering offerings in the stablecoin space that exists in 2026 compared to you know like 2018

Ran Goldi  17:36  

So so let's go under the hood, Alex, because again you are a senior staff engineer, and I do want to ask you some hard questions. I mean, first of all, you let's talk about the stack because I think that it's important to to hear about the story behind this product. Again, that you chose to build this with, you know, with with Bridge, which which is one of the great company, obviously part of Stripe, with with Ramp explicitly, I think you know you you wanted to use third parties as well in part of this process. What was your thinking through this? Obviously, Bridge Stripe very reputable, but you had to bring to a company that is very I guess you know they're used to working with bank APIs, and they're used to work with an API that if they move money, they think that it's it's all great, right? That the bank will take care of it. Worm case scenario: there was a bad Swift message, it bounces, whatever. You know it's recoupable here on the blockchain. Did you have to explain to people that it's way more dangerous?

Alex Bazhenov  18:40  

Yeah, Goldie, that's a great question, and and I think you kind of like really hit the nail on the head. Where like Ramp is an organization that's I think six years old at this point, and has traditionally been on kind of like the fiat side of money movement and the fiat side of money storage. And obviously, moving money and storing money on chain is like quite a different beast. We like one of the reasons that we chose Bridge, and then we also work alongside Privy in that capacity, is because there is this kind of like abstraction layer that they provide that make it a lot easier for engineers or operators that are not necessarily even familiar with you know what a scanner link is to go in and understand like what code is doing, to go in and understand what is happening with transactions, kind of like under the hood. A big part of the job of the stablecoin platform team at Ramp is to abstract all that complexity away even further, right? So, I would say that like our guiding principles are anything related to, for example, contract addresses, anything related to RPC calls, anything related to reconciliation, on-chain reorgs, everything stablecoin specific like that sits kind of like solely within the the purview and the responsibility of the kind of like more crypto. Stablecoins team and every other part of Ramp, including our customers, right? Like, really don't have to think about like the fact that they are interacting with a blockchain or interacting with a with a banking system. Like the goal that we have and that we've pretty much built towards and are now expanding even more is that the like intent and the primitives that other systems should see, whether that's a clicking a button in the UI or that's maybe a ledger system interacting with a payment system, is I have asset A. I want to send it from location B to location C, right? And then everything else under the hood is abstracted. So the nice thing about that is there wasn't this enormous period of ramp up for the rest? No pun intended, for the rest of the engineering org. Do you use that a lot

Ran Goldi  20:47  

in your internal conversations? Like, oh, let's ramp up. Let's do the. Honestly, it is.

Alex Bazhenov  20:52  

It's a. It's a mind virus. I can't say it now without thinking it and then saying no pun intended. But it's one of the things that come with with with ramp,

Ran Goldi  21:02  

but wait wait before before we move forward. Sorry, go ahead, Simon.

Sy Taylor  21:05  

I was going to say days. ramp.com is is still one of my favorite websites. If you've not been there, I'm sure you guys see it far too often.

Ran Goldi  21:12  

Oh wow! I I want I want to ask one more question about under the hood. Maybe I'll ask some more. But like you guys are today, you run on the base chain, right? Coinbase's base chain for accounts, and you rolled out USDT, I think, across ETH and Solana and Plasma. But like, what's your actual chain selection? What would you like? What would make you add a chain? Maybe you know Tempo or my own Elmo chain that I'm trying to pump here on the chain, right? I'm trying to pump it here on the on the podcast. Or what would make you kill a chain, right?

Alex Bazhenov  21:46  

Great question. I I don't think I've thought about the second one in the the second part of your question in the context of Ramp. Although I certainly have my own answer to that that I can that I can give. In terms of what would make us add a chain, it 100% comes down to customer demand, right? Like, if we have customers that are begging for tempo or begging for Elmo chain, I think that it's kind of like a no-brainer for us to add that. And we are very close to our customers, especially customers that are part of this our stablecoin beta right now. Although at the time that this episode's airs, I think we will have officially gone into GA, and then yeah, in terms of what would make us kill a chain, I don't know. Maybe maybe you can ask me on on Twitter, and then I can give an unofficial answer to that. I

Sy Taylor  22:35  

love it. Yeah, no, your your Twitter account is is exceptional, Alex. I will say I'm going to switch back to Andrew for a second, though, because I'm reflecting a little bit on what Alex has said there. I interviewed Eric 18 months ago, two years ago, on the FinTech Brain Food podcast, and he described this sort of like do what you're great at and let the third parties like with great APIs do that. And that's a model I think Alex was talking about there. But on the the business model side, it really is save time, save money, and a couple of things in the way you've launched this seem to be about that. But I'm I'm worried, I'm thinking about how are you guys making money. So you launched with no on ramp, off ramp, or gas fees. So you're covering all the gas. So like, how how are you how are you monetizing here? Is this float? Is it interchange? Like, is this making money for you? Like, what does that look like?

Andrew Chapello  23:25  

Yeah, I mean, I think we tend to think of monetization at like a whole company level, and and tend to think of the long term first. And like my philosophy is more if we're able to get customers that care about this and able to solve customer problems in a more elegant way. Long term, the monetization is is going to get figured out. I also believe that like on and off ramping fees in general are going to go to zero. You know, with with the like OUSD launch, one of the things that we were particularly happy about was that minting fees were at zero, and we've seen other providers do that as well. And so we want to monetize on ways that are

Ran Goldi  24:06  

whatever providers. We still haven't got that memo, by the way. Okay,

Andrew Chapello  24:12  

I mean it's going to evolve over time. But like you know, if we can monetize it in ways that are well aligned with our customers' interests, like you know we've got a really powerful software solution in Ramp Plus that we continue to add great functionality to. If we can make that you know the best software system in the world for time and travel expense for AP, you know then we don't need to monetize as much in other places. Also, like a fundamental principle that we have like if you have $1 that that should be $1 no matter where you go. Like it's we don't want customers to feel like we're all the time as they're moving their money around and and want to make sure that we sort of maintain that principle throughout.

Sy Taylor  24:54  

Yeah, I've heard you say that a few times in replies to me on X. Like $1 is $1 and I think that's sort of. Catching on now is a meme. Like it's a lot of places have started to do it. Revolut started to do it, just like honoring $1 and you guys take the pain away for customers. But as we see trust charters coming, and we see banks like Standard Chartered looking to do minting and redeeming, that starts to become something. To your point, that the on and off ramping starts to really blur, and you've got large financial institutions that can help you in that in that system. But as you're talking there about thinking about the long term, like I'm thinking, like isn't card interchange your core revenue engine here? Like, where's the where's this going to be going if you're disrupting yourself? Like, are you thinking about maybe yield, I know you have rewards in there. Is is that going to be the mechanism?

Andrew Chapello  25:45  

Yeah, I think you know stablecoin back cards. You know we're doing that with Stripe, and that's been been a successful product for us. And we think that there's there's still monetization to be had there. Frankly, I think there's a lot of room to run in like AI and automation on top of the core money movement for us, and and you've seen us come out with some new products around that. You can expect a lot more in that space. So I think we'll continue to lean in there. The different yield opportunities are pretty interesting, and and don't quite know where we're going to land on those yet, but expect that that there will be continued innovation at that frontier, and there might be opportunities there. Alex, anything you want to add? You got a good perspective here.

Alex Bazhenov  26:31  

I mean, I think at the end of the day, it's all about utility, right? Like like ramps.

Sy Taylor  26:36  

Crazy sentence. This is you know what that we we call tokenized. I'm kidding. I'm kidding.

Alex Bazhenov  26:43  

But yeah, like like Ramp's product offering is about utility to businesses. We can try to find like niches where, like Andrew said, we can kind of like take money out of you know like little slivers and and participate in races to the bottom. I think that stablecoins are kind of like the next step forward in terms of the utility of money, and if we are not there building with it and building at the forefront of it, we're going to fall behind, and yeah, we'll we'll figure it out.

Ran Goldi  27:13  

Look, Alexandra, I'm sorry to tell you, but no one will blame Ramp at falling behind. Like you're literally now the company that's considered to be at the forefront of of all these solutions, whether it's how to implement AI as a company, and we'll talk about that soon, or how you've implemented stablecoins. So you you guys are doing an amazing job. I want to ask another maybe tough engineering question, if I can. Like again, engineering myself, and obviously we're at Fireblocks. We have a lot of engineering challenges because we need to move $8 trillion a year of people's money. What was the hardest engineering problem you guys had, Alex, putting this tech in? And I and I and I sort of like have three guesses or three options in my head. It's either reconciliation hell between ERP cash equivalent accounts to stablecoin accounts that your team had to do. I don't know. Maybe you use your own month close, or was it like address risk for wrong networks that causes sometimes like 60% drop in funnels, or was it the fact that you sort of like had to staple together 24/7 rail and rail with banking partners that closes at five, and you had to bridge that? What was the toughest problem for you guys?

Alex Bazhenov  28:35  

Yeah. So first of all, I want to give a shout out to our design team because when I was coming into Ramp, I thought that address risk would be kind of like the bane of of our existing, but it actually has turned out to be like like not that significant yet, at least compared to what I've seen at other custodians in the past. I think it's definitely the last part, and I think I think even more so than that, it's marrying the like constraints and abilities of a fiat ledger that is six years old at this point with the realities of a blockchain ledger that is available 24/7 that can hold multiple different types of assets that can't go negative in balance. For example, that was an interesting thing for us to tackle because we can have negative balances in our fiat ledger. You can owe ramp money, and that can be a negative balance. And then obviously, you can't have a negative balance of USDC or USDT on chain. And then doing that in a way that let us ship this product quickly and not in six months or nine months or something like that. I think that there are a lot of interesting like trade-offs between having a traditional double-entry ledger and having a like a portion of the ledger represented on the blockchain. We're still working towards being able to marry everything internally in an elegant way. Like right now, you can hold. Multiple different types of stablecoins on Ramp. You can transact in different multiple different types of stablecoins on Ramp, but it's not. I wouldn't call the internal solution necessarily like elegant. I would call it like pragmatic, and we've done it. But if we were to rip everything like apart and build it from the ground up, we would of course build it very differently. But that's that's been the most challenging, but also I think the most fun thing to work on because it's very much a problem space that you don't get to work on as a crypto engineer working at a Web three company where everything sits on the blockchain by default.

Sy Taylor  30:33  

Yeah, no, it's edge cases all the way down in fiat, isn't it, Andrew? It's it's it's kind of like mismatches and timings and payments against accounts and one weird random quotes. There's a good friend of mine, Wade Arnold, who you may know. Andrew once told me that payments are actually really easy. It's like an email; you just send a message. It's the edge cases that are hard, and it's 99.9% edge cases. So have fun. Like it's it's engineering problem one on one. So true. This is a perfect opportunity for me to just take a quick pause while we thank our sponsors. Tokenized is also sponsored by Fireblocks. Fireblocks is the stablecoin infrastructure of choice for global businesses from Visa to WalPay to Bridge to Revolut, with over $100 billion in monthly stablecoin volume, Fireblocks powers stablecoin strategies at scale with infrastructure that enables PSPs, fintechs, remitters, and banks to issue, move, hold, and manage stablecoins. It's all done securely at scale with secure built-in compliance. With Fireblocks. you get complete control to build your own stablecoin orchestration layer, create payment accounts, manage liquidity, and access on and off ramps in over 60 currencies. Makes it easier for you to build and scale and expand your business globally. Learn more at Fireblocks.com. Thank you so much to our sponsors. I'm going to come back in. We were alluding to AI there. We made it like half an hour without really going all the way in. But but yes, to the point. I wrote an entire report on like how companies are building operating models with AI, and like Ramp was one of my star pupils because frankly, you just publish so much about how you're doing stuff internally. There might be companies further ahead. They just don't talk about it as much. So if that's like employee attraction hiring thing, really smart move. Just publish everything on X, and it's really useful. But you went from agents for controllers in July of last year to AP agents in October to full procurement agent fleet and now applied AI solutions with forward deployed engineers into your customers. So was the split here between agents that act like really value add autonomy versus like glorified copilots with approval cues, because, like, frankly, I speak to so many people in enterprise. They're like, they see all of the press that comes out from a company like Ramp, and they're like, "Yeah, I have this copilot thing too, and it's rubbish. It's Microsoft Clippy. Like, take me inside that disparity.

Andrew Chapello  33:17  

Yeah, I think it's it's been really fascinating to see the market evolve over the last couple of years in this space, and the the areas of utility start to make themselves clear. And so, you know, fundamentally, we believe that that the adoption of AI is the biggest technology shift that any of us have seen in our lifetimes. And so, we want to make sure that we can make this technology available to our customers and use it as we keep saying to save them time and money. And so everything starts from that. And like fundamentally, we think that AI can help save our customers time in automating their processes and save them money in making their teams more efficient. And we think that finance is kind of an ideal space for this because a lot of it is is so process driven, and in a way that you couldn't necessarily teach a deterministic system. You know, here are the like 125 steps of my month end close, and if this doesn't happen, then go down this other process tree. It's a lot easier to teach an LLM how to how to handle those processes, and LLM with the right skills and the right tools built in. And so, whether it's you know the the agent around your your company's expense policy or around sort of reviewing your bills as they come in, or stack as I mentioned, which is primarily focused toward accounting firms and automating those processes, it all comes back to how can we make these processes that have been sort of process driven and over over a long period of time easier. How can we help your team focus on? The more value added work, rather than the repetitive tasks over time. That's where we've seen like really big gains, and that's that's externally. And you know, you mentioned a lot of the work we've done and published. I think internally we are really focused on how we can use AI. What we say a little bit tongue in cheek is is we want to use to become the most productive company in the world, and that's you know a grandiose mission. But it really is something that we live and breathe internally. Is how can we use AI to write better software to make our teams more efficient, so that we can spend more time thinking about our customers.

Ran Goldi  35:37  

And and and again, we're we're it's not just a normal company we're talking to about AI. It's really because you know, and I'll give I'll just give an example. I have a friend that started working at Ramp, like literally, I think it was like three weeks ago. No, sorry, three months ago, right? You know, no names. His name is Zif Ahmed, great guy, and he said that like what's amazing at Ramp is that you know how you need to figure out what you need to do throughout the week and who you know who what's your most important projects and what are your tasks and how how are you moving forward with your KPIs and your goals. He said like at Ramp it's just a prompt. It's one prompt. It's not even like a series of prompts, it's like you have a whole operating system basically at the fingertips of your hands. I don't know if it's true. I believed it. I now want to work at Ramp, but but seriously, on on how you guys run your own tools, I can only assume that Ramp's own finance team again runs on this layer, right? They they run on your accounts payable agents. They run with your you know your policy agent, right? That that now catches 7x more fraud cases, right? And and again, it handles capital planning. It handles various analysis and so forth. But I think the real interesting question to you folks is how did dog fooding really change the product, and I have to ask this additional follow up of where does stablecoin fit into this?

Sy Taylor  37:11  

Andrew, you want to take a swing at that? So this is like your dog fooding this stuff. Has it changed how you work? And you know, is the finance team looking at stablecoin seriously for its own operations.

Andrew Chapello  37:23  

Yeah, you know, I think like our finance team is is our first customer for for all of our products, and and is is on the front lines and constantly giving us feedback, constantly telling us what needs to be better about our product. And so, you know, the first version of anything that we have, they're they're what one of the customers that we roll this out to, that we ask questions of, and that we use to stress test it, and they are not shy about giving us critical feedback on any of it. And you know, in some cases, the first version of a product doesn't quite meet the mark, and we need to to iterate. And like our imperative is, we need to do that as fast as possible, we need to take take feedback and get it into the product, and that's why we like have invested in our coding agent and inspect so heavily. You can take a bug that a customer reported and in Slack have a PR ready in a couple of minutes. It's it's really been been extraordinary to to build with them. In getting into your question about how we're using stablecoins internally, you know, when you think about Ramp, it takes a lot of capital to operate our business, especially like the card business where we are making loans to customers in different geographies. Money needs to be in a certain place at a certain time, and and that's really the efficiency that we see with stablecoins is being able to hold money where we need it, and then being able to deploy it to where it needs to go at a given time. So you know, for for example, if we have customers spending on cards in Europe, and we need to sort of fund a lending facility in Europe on a weekend or on a U.S. bank holiday, and that money is in the U.S. The easiest way to get into Europe is in stablecoins, and so like finding these these pinch points is is really like the core use case that we've been focused on with our team as far as stablecoins go.

Sy Taylor  39:18  

Alex, what did you have to build to make that happen? Like, really interested in what you've built on the back end with the finance team and what it's what it's felt like building this stuff. Yeah, so well, all right, I'm gonna say something controversial, which is that

Ran Goldi  39:32  

don't say agentic. We're not there yet. Don't we're not

Alex Bazhenov  39:35  

there yet. We're not there yet. Which is that faster payments are better payments, and wow, right? Like extreme controversy. And I think that when most people hear that, they think of like, okay, yeah, the customer is getting their payments, or the counterparty is getting their payments faster, the sender sending them faster. It's great. You have more like capital efficiency and whatnot. I think that there is a like a second tier to that, which is probably. Clear to like operators and people that actually work in money movement, but might not be clear to the like like the wider space, which is that if a payment is faster, that means that less things can go wrong, right? Whether that's in fiat, whether that's in stables, like yeah, you know, you might you might be sending money via Swift. It's not great that it takes three to four days, but guess what? There might be a data center outage, and your payment could get like lost, right? Or some kind of you know like banking API failure. All of those things incrementally, or all the probability of those things incrementally add up when you're a business that's processing you know like millions of payments, whether that's on behalf of third parties or that's on behalf of yourself for kind of like capital efficiency, rebalancing, sending money cross border. So a lot of the work that we did focused on just making payments faster, and that's something that stablecoins do. And we're we're getting efficiencies from that internally. We're getting efficiencies from that externally for our customers. And I think that yeah, without getting like too into like engineering details, the the bulk of the work that we have done was how do we make this faster? How does how can we measure that that decreases our operating costs? Whether that's time to resolve a ticket, number of tickets resolved, errors or edge cases that we don't run into anymore, and how can we scale that to kind of like ramps internal finance processes and to our external like third-party customer payments as well?

Sy Taylor  41:32  

It sure helps to have the ability to create a pull request in Slack almost immediately once you once you've lived that sort of customer complaint or internal customer problem to PR compression of timeline. It's it's hard to go back. I'm telling

Ran Goldi  41:47  

you, if if I hadn't had to move $8 trillion at Fireblocks, I would move to Ramp tomorrow. I

Alex Bazhenov  41:52  

will say that, like, yeah, I will say that, like, I think that working without the AI tooling that we have at Ramp in any capacity, whether you're on the sales team or on the engineering team or on the finance team or the the customer experience team, would be like working in molasses. I can't imagine going back to an organization, yeah, that doesn't have the level of tooling that we do. It would be like losing a limb.

Sy Taylor  42:20  

Tempo Giorgios, the CTO, built something similar called Centaur and open sourced it. And it's a really great tool. Not quite the full harness that you have with Glass in terms of context around a user. That bit really excites me, and I want to build that next. But this ability to just very quickly have high quality engineering that really understands your codebase, so that you can get a pull request up to speed. And I'm imagining, though, like speaking of speed, Ramp is all about speed, velocity of transactions, move money faster. Makes sense that you want payments to go around the world a bit faster. But another thing that might make payments go faster is if your customers all have agents that are making payments, or consumers have agents that are making payments, like Lord knows the card networks are trying to make this happen with Visa Intelligent Commerce, and I know you guys got involved with some of that. But like Andrew, how do you think about how risk works in this universe? KYC identity, like there's a lot of unsolved problems here. If we're going to have agents moving money faster, they can do so internationally with stablecoins, they can do so with a card now. How's this going to play out?

Andrew Chapello  43:25  

I think there's there's a couple primitives that we're that we believe agents are going to need to do commerce at a large scale. You're going to need to be able to specify what we like to think of as spend controls. You know, these are the merchants that you can spend at. Here are the amounts that you can spend. Things around velocity limits. Maybe you'll want to have in early stages human interventions at different stages and process. I think that's a core primitive belief. Yeah.

Sy Taylor  44:00  

Sorry to interrupt you, Andrew. I saw really interesting consumer the new consumer report. I'll send it to you. That was looking at like consumer preferences, and I imagine business preferences as well about that human in the loop thing, which is like people are quite happy to take medical advice from an AI, but they want to get in the way of it if it's going to make a payment. Like you can, you can tell me that this, you know, that I'm not dying when I really am, but don't tell me to move, don't move my money on my behalf, like that in the loop thing. I think is going to be really key. Sorry to jump in, but you you excite it.

Andrew Chapello  44:31  

I think that it's built up off of years of fraud and really scary stories that you hear both in headlines and from friends about like where can this go wrong, like can can an agent drain my bank account? Is like a core fear that we need to overcome before this really becomes mainstream. And so yeah, we think about spend controls. We also think about like the different purchasing. Mechanisms that you use, and so like there have been a couple of different theories that that that people in this space have had. You know, I think stablecoins are really interesting in that you can transact at really high volume and really low cost. Cards are also really interesting because they already have a lot of the fraud protections built in, but like what what we found with cars, at least in traditional checkout, is the protections that have already been built around robots and around sort of automated protections are stopping a lot of the agentic commerce from happening in those times. Well, because

Sy Taylor  45:39  

it just looks like a script by a fraudster, it looks like a bad bot. Yeah,

Andrew Chapello  45:44  

certainly. And so I think that's a problem to overcome. But we still fundamentally believe that there's a place for for cards to give agents cards to spend on on both consumer and businesses' behalf as well as stablecoins where that's relevant. Where like, and I think stablecoin opens a whole new method of commerce. Like at least in B 2b like the the traditional model is like, oh, I I select a vendor, I sign a contract, I have some sort of minimum that I'm going to hit, and and then after all this work is done, then I can actually go and work with that vendor. I think the big opportunity is like, can you put that? Like, can you can you like can you solve the problem that you're trying to solve with contract, which is like trust, like you will pay for this with with stablecoins, where you know the payment is final and it's instant and it can be done any amount. I'm hopeful that a new business model will emerge around that.

Ran Goldi  46:39  

Amazing. I I think that look. By the way, Simon can talk to you about this identity and compliance aspects for years. Don't forget, he came from Sardine. This guy knows too much, so he he can definitely talk about this for for ages. I we're just about to get to the to the end of this of this episode. Unfortunately, we we could have continued this forever, I think. But there's so much to ask before I ask for an official role at Ramp. But I I want to ask you I guess you know the question we ask everyone and Simon maybe you want to add to that like 20 years from now, right, Andrew and Samala? You ask Alex something surprisingly different, but Andrew, 20 years from now, are stable coins just a gimmick that a company that was doing AI platform and for finance used, or is it a fact in our financial system?

Andrew Chapello  47:37  

It's going to be bigger than Swift.

Ran Goldi  47:39  

Oh wow! Okay, that's a bold statement. Swift just launched 24/7 tokenized deposit payments with 17 global banks. So, but I agree with you.

Sy Taylor  47:50  

Yeah, no, the Swift ledger does orchestration. It doesn't do settlement. So it's a it's a very different animal. That's a it's it's a useful thing, but it is not the same thing as settlement. And I think we just got our episode opening segment there. It's going to be bigger than Swift. That was a that was a soundbite and a half, Andrew. Alex, I wanted to ask a question for the operators listing, which is like, what have you learned building over the last 18 months with stablecoins at a business your scale that you wish you knew at the beginning that people should take home as like the one bit of useful engineering advice from this show,

Alex Bazhenov  48:23  

I think that with proper preparation and with the like correct level of abstraction, which is a difficult level to achieve, you can put stablecoin, stablecoin payments, stablecoin settlement into a. I don't know if legacy is an offensive term, but a legacy payments business or fiat business, and it's a lot less scary than it seems.

Sy Taylor  48:49  

That's it right there. You can do this, people. And if you can't, then tempo dot xyz forward slash advisory, and me and a team of FDEs will come help you do it. I love it. I

Speaker 1  49:00  

love it.

Sy Taylor  49:01  

It's it's it's what we're here for. Listen to great people, and then soak in the knowledge, and then go do advisory. It's it's it's a good business model. But look, Alex, Andrew, in all seriousness, you guys are at the forefront building this stuff. You're the reason I do this show. You're the reason I run fintech nerdcon the event, and you guys will be speaking there on the 19th and 20th of November, and in San Diego. I strongly recommend people get a ticket to to hear more of what you guys have got to say. I want to thank you guys so much, and I also want to find ask you, Andrew, start with you. If people want to get in touch with you or find out more about what you're building at Ramp, where do they go to do that?

Andrew Chapello  49:40  

Yeah, I mean ramp.com for to find out what we're doing at ramp. But I'm on Twitter. It's just my last name. Hello, I'm Ramp on Twitter.

Sy Taylor  49:48  

Yeah, good, good handle, guy. Alex, how about you?

Alex Bazhenov  49:51  

I am on Twitter as well. I am ZeroX Dataless, and I'm not going to try to spell that. But if you want to find me, you probably can.

Ran Goldi  49:59  

You'll see it on our. Show notes, and you can find me Goldie at X on Rand Goldie and LinkedIn at Rangoldie or fireblocks.com/simon Taylor Advisory Price Pro vocationals.

Sy Taylor  50:11  

That's what it is. That's what I always needed a Goldie out there selling me. Look, you'll find me at S Y Taylor on all of the socials, screaming into the void at fintechbrainfood.com and of course at tempo.xyz/advisory. But not least, you'll find a lot more of me and this show if you subscribe to it and you spam all of your friends to do so too. I'm sure Alex and Andrew will agree we're putting out good stuff with great people and it's more than worth it. Thank you very much for your time and catch you next time.