# Mural has raised $28 million on bet that every business becomes a fintech

Stableminded, a Stabledash show. Nov 19, 2026. Sinclair Toffa.
https://live.stabledash.com/shows/stableminded/episodes/every-company-and-person-is-becoming-a-fintech-few-are-ready-PWuaZmDof7k

Sinclair Toffa says AI agents will change payments the way mobile did, and that they should get dedicated accounts rather than open access to people's banks.

Mural Pay has raised $28 million to date, its founder Sinclair Toffa said in a recent interview. The company's pitch rests on one thesis: every business and every individual is going to become a fintech. Toffa says becoming a fintech means bringing embedded financial software into a business so it can launch financial products and own the end user experience.

## Everyone is a fintech

Sinclair Toffa names Uber and Starbucks as fintechs, and says Anthropic is becoming one too. He points to Opera, which he calls the fourth largest browser in the world and a customer, as a company that became a fintech after noticing people in emerging markets use the browser every day.

At Mural Pay, one account object abstracts away both stablecoin and fiat across different currencies. Toffa says the company sends forward-deployed engineers in person to Mexico City, Buenos Aires and Bogota, with a goal of reaching a first production payment two weeks after signing.

## Lessons from Latin America

Mural Pay worked with Deal on a stablecoin product for Latin American employees. Sinclair Toffa says that for payroll to be legal, money has to flow from Deal bank accounts, so paying into a named employee account first is an important compliance step. Mural is also working with XXL to turn a network of bodegas into mini fintechs with global accounts.

"No one cares about stablecoins, really," Toffa says. Users in Argentina care about getting money into neobank platforms such as Dollar App.

## Agents need their own accounts

Sinclair Toffa says connecting an agent to Plaid gives it full read-write access to all of a person's bank accounts, so the agent can start spending. He wants dedicated global accounts instead, where money is allocated and the agent pays out by bank transfer or PIX. He predicts mobile apps will be wiped out within five to eight years.

## Build or buy

Sinclair Toffa sorts companies into primary, secondary and tertiary fintechs. He says primaries, such as Ramp, are best positioned to capture the agentic wave, but many try to build blockchain infrastructure in-house, find it more complex than assumed and keep pushing back launches. He advises them to whiteboard the architecture first, and tells tertiary companies to partner so they can focus on distribution.

Toffa's case for Mural Pay comes down to one bet: as more companies hold money and agents start spending it, they will want compliant, programmable accounts they can plug in, not infrastructure they build from scratch. He says AI-native tertiary companies are well positioned to build on such accounts quickly.

## Key takeaways

1. Mural Pay has raised $28 million to date on the thesis that every business and every individual is going to become a fintech.
2. Sinclair Toffa calls connecting AI agents to Plaid probably the worst thing you could do, and argues for dedicated accounts that agents spend from.
3. He predicts that all mobile apps as we know them will die out within the next five to eight years as agents become the way people express intent.

## Chapters

- 0:00 Cold open and sponsor read before Sinclair returns
- 2:19 Mural raised $28 million on thesis every business becomes fintech
- 8:14 Palantir taught him to build ontologies of core business objects
- 12:41 Mural's API treats stablecoin and fiat as one account
- 14:14 Forward-deployed engineers aim to get customers live in two weeks
- 18:06 Deal payroll and bodega networks show the model in practice
- 27:28 Agents replace mobile apps as the way humans express intent
- 31:23 Give agents dedicated accounts, not Plaid access to every bank account
- 36:37 Fintech centrality sorts companies into primary, secondary and tertiary
- 41:23 Whiteboard the build-versus-buy decision before stitching providers together

## Transcript

**Sinclair Toffa** (0:00)

No one cares about stablecoins, really. The way I explain Mural now is we are building the compliant system of record for businesses to launch financial products. And it was at Palantir where I realized the incredible fragmentation of this world. It was our job to build the ontology, which is what are the core business objects that shape your company. In order to build great ontology, you have to be an excellent architect. Let's look at the architecture of things from the past.

And try to draw parallels to understand how this whole agentic wave is going to play out. Agents are the new way we express intent as humans. The agent is this self-contained fintech. People are connecting their agents to their plaid, which is probably the worst thing you could do. There's this word that I love called apophenia, the ability to find patterns amongst disjoint things that you normally wouldn't connect together. What is your core ontology? What are your core business objects? What do you want to own?

**Drew Rogers** (1:04)

Stableminded is proudly presented by Latitude, your go-to partner for simplifying global payments. I've interviewed hundreds of teams building in modern money. No one has the enterprise track record of the Latitude team. With Latitude, you can move money to any continent in the world in a regulated, compliant, and secure way. Build with global ramp and payout infrastructure now available in over 40 countries. Single API to build any funds flow you can imagine.

If you want an intro, you can pay me on X or go to latitude.xyz to learn more. Sinclair, we're back.

**Sinclair Toffa** (1:35)

Third. Third time around.

**Drew Rogers** (1:38)

First time we're doing this. One of the biggest joys of doing this is following people through the journey of like building. And it's been so cool to see, I think it was actually this month, Two years ago, where we had our first podcast conversation in season two. So like episode probably six. We're now in the sixties, I think. So thanks for coming back. Thanks for having me back. Why don't we start with the centering this conversation around like the architecture mindset of how you build lasting things.

Like I would love for that to be a through line for this conversation. But why don't we start with Definitely the recent news, a little bit about Mural for people who don't know you yet. And then we could talk about how to build things that last.

**Sinclair Toffa** (2:19)

Beautiful. Well, thanks for having me back. And it's been a very exciting past couple of weeks for us, past couple of months. We announced a recent fundraise, so we've raised $28 million to date. And we've been in full growth expansion mode with the team, with the customers. And I'm excited to talk more a little about How to become a world-class architect and why I believe that's very important in terms of how to win in today's day and age.

The way I explain Mural now is we are building the compliant system of record for businesses to launch financial products. These financial products allow you to store money, move money, and grow money. Very simple. And it's rooted in the thesis that every business and every individual is going to become a fintech. And what I mean by that is it reminds me a lot of when I was at Palantir as a floor-deployed engineer. We were doing the transition from on-prem data storage to the cloud.

And it was this transition of become a cloud company, become a data company, right? And you had two types of companies. You had companies that realized, okay, we're going to migrate to the cloud. and become a data company, and then others who said, we're not going to do

**Drew Rogers** (3:31)

it.

**Sinclair Toffa** (3:32)

We already have all of these. And the ones who became cloud-based companies, data companies, were able to create all these data applications to make their business more efficient.

**Drew Rogers** (3:42)

And now every company is a data company.

**Sinclair Toffa** (3:45)

And so I believe we're in a very similar phase right now where every business, every individual is going to become a fintech. And you have two types, right? You have the ones who are embracing that now and are becoming fintechs now. And that's going to be a go-to-market, a business model, a distribution advantage. And then you have those who are not doing that now and are only going to realize this in the coming years. And the ones who adopt that now will have a competitive edge.

**Drew Rogers** (4:11)

Yeah. And so when you say every companies like which I think I'm following every company is a data company, meaning like most industries and most companies are utilizing or building or leveraging within core systems. What at that time would have felt like kind of cloud related systems to run and operate the company like every company is using something like a cloud served inbox or notion or drive or like we now have this state of business that is cloud native by default is kind of what you're saying.

And that's going to be, in your view, how fintech or financial systems evolve over time. That's correct.

**Sinclair Toffa** (4:47)

The world is becoming more financialized. People are storing money in all these different platforms, all these different businesses. Every business generates money, hopefully. And as they scale up, they'll want ways to further retain their customers, further monetize their customers. And one of the best ways of doing that is enabling your users, your customers, to keep their money with you. And so we're noticing that with Uber is a fintech, Starbucks is a fintech.

We have a lot of food delivery companies that have been approaching us who are in the process of becoming fintechs. We've talked about Opera, fourth largest browser in the world, who's one of our customers. They have become a fintech because they notice that people use the browser every day in all these different emerging markets. So why not allow them to keep money in there and pay for their bills directly from their browser?

You can imagine a world where, you know, even a business like Spotify, for instance, could become a fintech or you're noticing that companies like Anthropic, right, are becoming a fintech as well. And so money is getting stored in all of these different places.

**Drew Rogers** (5:50)

When we say fintech, that's just kind of we could start mentally imagining a world where Drew Rogers, Sinclair Toffa, Zach Fowler, Incline, Sam Breckenridge, Stabledash Global, Stabledash Live, Stabledash Live.

**Drew Rogers** (6:10)

For whatever the use case might be, I'm definitely a believer of this. It sounds like you are too, of like wherever the value is originating, that value will hopefully retain there so that the companies like Spotify or whatever that might be, the people who are building the value, they also retain the economics around the value itself.

**Sinclair Toffa** (6:29)

That's exactly it. That's exactly it. And it's important to define what it means to become a fintech, which basically means bringing in embedded financial software into your business so that you can launch financial products and you own that end user experience. So you can launch a wallet, you can launch cross-border payments, you can launch a rewards yield program product, for instance.

The reason why compliance system of record is so important is because the fragmentation of all of these different financial tools, all of these different financial systems is going to keep growing exponentially. Every country has its own systems, own compliance, own regulatory posture. All of these different systems are being developed in unique ways. You have blockchain-based systems. You have fiat-based systems. And that fragmentation is getting very complex. And so we want to absorb that complexity.

We want to take on that complexity in the compliance that's being that system of record and building that trust is really, really

**Drew Rogers** (7:32)

important. Even the word fragmentation, bringing in a little bit of this architectural way of thinking, that word fragmentation, if I'm building something, fragmentation at the foundation is not good. We don't want that. Bring us in a little bit to, and feel free to go wherever you want to go with this. Like I know Palantir has been super formative with like how you think about architecture. I think the first time I heard the word ontology was from you and like what that means.

The way that you have then taken that experience and thought about architecture with Mural, like even the naming and branding, like bring us a little bit into how you've now built this architecture and how you think about modern architecture.

**Sinclair Toffa** (8:14)

So to start, I would look at just how I grew up. Growing up, I was just obsessed with data. I was obsessed with data, I was obsessed with system building, and I was obsessed with creating ways to self-improve. And so what that meant is from the age of around like 14, 15, I started tracking every single expense that I ever made manually. Had a sheet, then an app. I still do it to this day, so it's been around 15 years of doing that.

And just tracking everything manually, building different systems to understand the patterns in my personal life. And then that translated over into when I joined Palantir. I felt like Palantir was just this amazing opportunity to understand how the world works from a data standpoint and how do companies store their data and how do they make use of it. And it was at Palantir where I realized the incredible fragmentation of this world. And people don't realize that.

You see these massive companies out there That are powering your food delivery, your mail delivery, cybersecurity, auto manufacturing, aircraft manufacturing, telecom, for instance. And they all had hundreds, thousands of different systems that had data in them. And so it was our job to build the ontology, which is, what are the core business objects that shape your company? And then how can we start off with the business objects, make sure that everyone's aligned on them,

**Drew Rogers** (9:40)

How do you define business objects?

**Sinclair Toffa** (9:41)

So business object is, and this is where we start getting into the architecture. So if you think of your company as a building, right? And the building has core systems in it, or you could even think of your business as a restaurant, for instance, right? And there's core operations being done, right? You might have like, let's take a

**Drew Rogers** (10:00)

I'm obsessed with the restaurant metaphor.

**Sinclair Toffa** (10:03)

We'll take the restaurant one. You have a customer object that represents a customer. You have an employee object or even like a chef object. You then have a transaction object. So you come up with these core objects that represent your business and it's really important to involve everyone from leadership to understand what those objects are. Especially from different teams, from compliance, from legal, from engineering.

Then once you come up with those core objects, you determine the relationships in between those objects. So how is the customer object connected to the transaction object, which is connected to the chef object, right? Once you set up the objects and the relationships, which makes up your ontology, then you can back into building out those data pipelines and understanding, okay, what is the data we need to collect and how do we clean it? And how do you create those pipelines to build a very clean, simple ontology?

And the clean, simple ontology is really important because from that simplicity, you can create very complex systems and you can grow quickly. The problem is when you don't have a structured ontology, you're starting off from all of these messy data tables.

**Drew Rogers** (11:11)

Yeah. So in some ways, my first job out of college was Square. So I was selling to restaurants in many ways. So Square as a company, Toast, it's kind of like a verticalized ontology software that they've just built and deployed for that specific use case. But that's a way of thinking and being that you would apply to any industry, like any-

**Sinclair Toffa** (11:31)

Everything.

**Drew Rogers** (11:31)

Yeah.

**Sinclair Toffa** (11:32)

Everything. And so in order to build great ontology, you have to be an excellent architect. And to be an excellent architect, it's important to notice a lot of the different structural patterns in the world. There's this word that I love called apophenia. And apophenia is the ability to find patterns amongst disjoint things that you normally wouldn't connect together.

**Drew Rogers** (11:56)

Can you give an example of one of those?

**Sinclair Toffa** (11:57)

An example could be like mycelium network and SWIFT and money movement systems, for instance, and international systems. So an example would be you look at the mycelium networks of how trees and forests communicate to each other, these fungal networks, and they're distributing messages in real time. And then you can compare that architecture of how that works to the internet or to how money should move. And we believe that money movement will be similar to mycelium network information transfer.

And so that apophenia, the more we can do that, The more we can become a better architect, the more we can find those patterns. When you become a better architect, you can then create a stronger ontology, a very simple ontology. And that is the exact way we build the mural infrastructure, the mural API. And so when you look at our mural API, we have a very simple, powerful ontology.

Which is rooted in all of those learnings from Mural, from Palantir, from building complex money movement, money storage systems, and data systems. It makes it really easy for our partners, our customers, to then build on top of that. And so you'll notice it's even small things where some other companies might have a distinction between a crypto rail and a fiat rail. And for us, we don't draw a separation between the two. An account.

**Drew Rogers** (13:17)

That's an object.

**Sinclair Toffa** (13:18)

An account is an object, exactly. Some other APIs out there might distinguish between a crypto account, stablecoin account, and a fiat account. So then, okay, you have two separate objects. So when you're building on top of that, you're now managing two separate objects. And then as you build on top of that, it becomes more complex. For us, it's one account that abstracts away both stablecoin and fiat and abstracts away across different currencies.

So you apply that concept to all of the different objects in our ontology, which makes it easier for you as a developer, as a builder, to launch new products and to fit our ontology to your ontology. And that's something that Palantir did really well, was really good at understanding businesses' ontologies and helping them develop that. It was very much so a collaborative educational process. Tell us about your business. How does it work? Let's see how we can do this together. And we do the same thing.

**Drew Rogers** (14:14)

Did Palantir originate the forward deployed engineer model? Did they come up with that?

**Sinclair Toffa** (14:20)

Palantir came up with that.

**Drew Rogers** (14:21)

Can you speak a little bit to the, like for those who don't know what that is, Why that's so impactful and how you've applied that way of building to Mural and how you actually like might need something like that to effectively architect meaningful things at scale. What that practically looks like right now for you guys.

**Sinclair Toffa** (14:41)

So when Palantir first came up with the concept of let's take engineers and deploy them with the customer, everyone thought that was nuts.

**Drew Rogers** (14:49)

When was that roughly?

**Sinclair Toffa** (14:51)

I mean, this was early Palantir days, like early 2000s.

**Drew Rogers** (14:55)

Oh, wow.

**Sinclair Toffa** (14:55)

This was early 2000s when Palantir was getting up and running. And so you'd take very smart engineers from the team, and then you would deploy them with the customer and say, OK, figure out the use case, solve their problems.

And at the time that was very novel and very different because typically the engineers would be coding and not customer facing and then you'd have sales or solution architects that would deploy with the customer and instead here we'd take the engineers, put them with the customers and At the beginning, there was a lot of learning experiences. You know, when I first joined, it was very much so we're looking to improve this process, right? How do we correctly do it?

And what was really unique about it is when you put engineers directly on the front line like that, you close the gap, the signal gap between the customer's pain points and where they want to be and what your product offers. And so you close that gap by decreasing the number of communication steps between salesperson to a product person to an engineer. Instead, you have the engineer feel the pain firsthand of the customer. And that closes the gap.

And then what's really important about the forward-deployed engineering model, it is not about building a bespoke solution for every single customer. That's a consulting business. It's very much so about, OK, how can we understand their pain points? And how can we see how we can best deploy the API into the business? And how can we make sure that they are doing this implementation to make more money or to decrease costs, right? And launching new products, expanding to new geographies.

And then for any gaps that we might have, the forward deployed engineer is actually going to ship code. Drew Rogers, Sinclair Toffa, Zach Fowler, Stabledash Global, Stabledash Live, Stabledash Live, Stabledash Live,

**Sinclair Toffa** (17:02)

Learn, abstract, redeploy, right? It's that loop. And that is what Palantir did. That's what Palantir did with the Foundry software, all the different softwares that Palantir built. And that is the exact model that we're using at Mural as well, as we have four deployed engineers on the team. They're deploying in person in Mexico City, in Buenos Aires, in Bogota, around the world. And they're understanding the exact pain points.

And then the goal is we're going to get you up and running from the moment you sign with us to first production payment in two weeks.

**Drew Rogers** (17:35)

That's the commitment.

**Sinclair Toffa** (17:37)

And I'm making the commitment to you right now as well. And that's the commitment we make to our customers is get live in two weeks. Compliance, integration, first production payment.

**Drew Rogers** (17:46)

Yeah, pick an example, whatever that might look like of like the last 12 months or so of like how that two week commitment or the localized forward deployed go to market, like how give an example of maybe a customer use case or problems that were faced or solved, like help make that a little more tangible.

**Sinclair Toffa** (18:06)

What have you been seeing? So I'll give two examples. Deal and this company called Axixel. So what Deal wanted to create when we first started talking was employee stablecoin product for Latin American employees. Allow Latin American employees to hold in stablecoins, you know, go from local currency into stablecoins. And when we deployed our four deployed engineering team with them, we uncovered a couple of really important constraints that were important to them and to us.

One notably, which was in order for payroll to be legal, money has to flow from deal bank accounts. So that intermediary step of paying into a named employee account is an important compliance consideration. And we uncovered that by doing the forward deployed motion with them and understanding, you know, why is that important? How do we architect that? How do we actually build that out?

**Drew Rogers** (19:02)

Is that a unique problem or is that something that you think is like, oh, that's like actually a felt pain globally?

**Sinclair Toffa** (19:07)

It's a felt pain globally. A lot of countries have laws on what is legal payroll and where it needs to land. And a lot of times that's tied to... It's

**Drew Rogers** (19:19)

very different. Like most when people hear like stablecoin sandwich, it's not the stablecoin sandwich. No, it's a reverse. It's like bread, bread, meat.

**Sinclair Toffa** (19:27)

Yeah, bread, bread, meat. Yeah, exactly. People think this is not the sandwich. Yeah,

**Drew Rogers** (19:32)

it's not a good sandwich, but it's compliant. Exactly. And it's interesting because a lot of people are like, the narrative the last couple of years has been like, you use stablecoins to move and settle really quickly and get from point A to point B. How can we actually end with a stablecoin but go through traditional hops to get there?

**Sinclair Toffa** (19:57)

In order for it to be compliant and

**Drew Rogers** (19:59)

legal. In order for it to be compliant and legal, that makes sense because you're going through these traditional rails. What do you say to the local in Argentina? What is the benefit of you end in stablecoins And so there's benefits not necessarily due to speed of settlement, but to actually having tokenized money now.

**Sinclair Toffa** (20:21)

So the benefit for the end user is not really stablecoins at

**Drew Rogers** (20:27)

all.

**Sinclair Toffa** (20:27)

No one cares about stablecoins, really. What they do care about, if you're in Argentina, is you're using Wallah, for instance, you're using Dollar App, you're using all of these new neobank platforms. And the quickest way for you to move money from your deal account, for instance, to your dollar app is by withdrawing through stablecoins. So it's more so a quick method of withdrawing it out or holding it as dollars. And this brings us back to the fragmentation component.

You have all of these new fintechs, all of these new fragmented tools, software, et cetera, out there. And so having a way to quickly connect in between these systems is really important. And what stablecoins are, stablecoins are built on top of blockchains. Blockchains are open databases. They have a structured data schema. And that data schema can be reused by everyone who's building a blockchain enabled app.

So that's very powerful because when you're connecting into a shared schema, that means there's less data integration that is needed. The reason why Palantir did so well is because everyone had different data schemas. You had data schemas all over the place, right? And Palantir realized that the security, the permissioning aspect was so important to unifying all those different data schemas. Same thing happening here.

You build a blockchain-enabled application, you're automatically connected to every other fintech, every other institution that has a blockchain enabled.

**Drew Rogers** (22:02)

Just to make sure I'm picking that up, I don't know if the listeners may be asking this too, when you kind of made that comparison, would Palantir in this example be almost at that like blockchain-like layer where they're kind of Like for the financial use case we're talking about, we have these like shared systems, data systems that are blockchains. Is Palantir almost like for their business, do you think about it as like a similar layer they kind of share in that data?

**Sinclair Toffa** (22:28)

Yeah, so I view the Palantir software sitting on top of all of these different data systems. Okay, so different. So different. Mural is this software that sits on top of all these financial systems, and we connect all of them. The unique difference is that we sit on top of fiat and blockchain systems. Blockchain systems are this unified data ontology, right? This unified ontology for how to move value.

And in the Palantir world, on the data side, you do have some standardized data formats, but every company has their own data format.

**Drew Rogers** (23:01)

Okay. I think I might have derailed you. Were you finished? Oh, yes. We're back to deal now.

**Sinclair Toffa** (23:07)

Yes. So that was on the deal side, right? Deploying our four deployed engineers and understanding The compliance and legal requirements around that. And we have this other customer, XXL, completely different use case, but same use of the underlying infrastructure. They're building these global accounts for, they have a network of bodegas in Latin America, network of stores.

**Drew Rogers** (23:29)

Yes, I can see their logo. Yes.

**Sinclair Toffa** (23:32)

The bodegas are these local stores.

**Drew Rogers** (23:34)

Mexico

**Sinclair Toffa** (23:34)

City, El Salvador, Colombia, Argentina. And when you go to your local bodega, it's like a family member, right? Because you're going back to your bodega every single day, every single week. And when you're going back to your bodega, you're paying for your bills, you're buying things, you might be sending money to loved ones, right? It becomes this trusted partner to you when you live locally like that.

And the premise is how to make all of these bodegas into mini fintechs, where now they have global accounts powered by Mural. And these global accounts have instant payments in and out between the US, Mexico, US, Argentina, US, Brazil, US, Colombia, for instance. And not only that, they're then creating these global accounts for the customers. So now you have a network of these bodegas and their most loyal customers who can now store accounts. The bodegas are now becoming a fintech, right?

And then these individuals have their own global accounts with the bodega and they can use it for real-time remittances, paying bills, everything in real

**Drew Rogers** (24:39)

time. Are these types of companies, like maybe this one specifically, like what was the before and after? Did they have like a mobile app that was just super slow and expensive or is this a completely new idea to them? Like are you actually going zero to one financially with them? I know like you think about MoneyGram, Western Union, like Yes. That kind of local bodegas often does serve as like this financial hub. Was this business already doing traditional finance movements?

**Sinclair Toffa** (25:05)

This is very much so old fashioned, brick and mortar, very little technology.

**Drew Rogers** (25:11)

OK.

**Sinclair Toffa** (25:11)

The main technology that people are using is WhatsApp. OK. You know, that's the main thing, right? And then they have, you know, local POS systems to accept money in, in local currency. That's the main technology that they have. And so this is very much so leapfrog advancement for them, where the forward deployed engineering motion is really important here because they're not fully financialized yet. You know, they haven't deployed financial software. There's a lot of on the ground work that needs to be done.

Doing this on the ground work is extremely rewarding because they have such deep loyalty with their customers and they have very strong retention. And so this is a way for them to increase that retention with their existing customers and just to increase that connectivity between the bodega and the US, for instance. And a lot of these bodegas, they're selling Amazon cards, Google Play cards, right? A lot of services, a lot of products from the US and vice versa.

A lot of them are getting remittances from their loved ones here in the US. And so that being able to keep in dollars Rather than needing to go in and out of local currency, allows them to save on the foreign exchange fees, allows them to save on all of the time spent coordinating, moving money back and forth. And those types of use cases I'm really excited about. These like massive networks, you know, they have tens of thousands of bodegas, millions of users, right? They're looking to connect all of them together.

**Drew Rogers** (26:36)

It's cool to think about the in-person tangible example that that is. It's like we can walk into one of these places and buy something. Can you now bring us a little bit into, in our first conversation, something I knew that you were already thinking about where agents digitally and on behalf of human beings are now in that position of traffic as commercial actors instead of, or on behalf of human beings walking into a bodega physically, like, Bring us into how you think about now where you have this architecture.

It's coming from a compliant first mindset. It's performant and can build the scale and how you maybe in the fundraising journey and like thinking about the future, like where are you guys at with seeing agents become a part of this system? Like what's that currently looking like?

**Sinclair Toffa** (27:28)

Part of this whole becoming world-class architect is let's look at the architecture of things from the past and try to draw parallels to understand how this whole agentic wave is going to play out by looking at the architecture of the past. One specific model that we look at, that I like to look at, is the shift to mobile. And the shift to mobile really kicked off in 2008, 2009 with the iPhone, with the App Store coming out, and you Mobile is really just a change in user experience.

And so we went from online websites as the core user experience for digital to buy things, sell things, etc. With you all

**Drew Rogers** (28:11)

the time. Yes. Not in a room in your house.

**Sinclair Toffa** (28:14)

Yes, exactly. To mobile. Where you could then buy things, sell things from mobile, right? And so you had this whole new wave of businesses that launched mobile first. You had Airbnb, Uber, etc. All of these complex businesses that we could have never imagined before. And so you had these mobile native businesses. And because the UX was that much quicker and easier, a lot of these mobile native businesses absolutely exploded.

And now we're entering a very similar shift with agentic and the way we can think about agents. Agents are a shift in UX and user experience. Agents are the new way we express intent as humans. So rather than clicking a bunch of buttons, downloading hundreds of mobile apps in order to Have one mobile app for ordering a car, one mobile app for ordering food, another mobile app for Google searches, another mobile app for paying someone out in Argentina. You now have agents.

What agents allow you to do is it allows you to chat over WhatsApp, chat over Muse, chat over Grok, chat over Instinct, and do all these functions in a much quicker and more seamless user experience. And so the reason why I am so excited about agentic and fintech is because agents are a new user experience innovation to make it that much easier for you, the consumer, to purchase things, send money, pay for a contractor, for instance. And so agents are going to be the new way to express intent.

And my hot take right now, but I don't believe it'll be a hot take, is all the mobile apps as we know it on our phone are going to die out within the next five to eight years. All of them are going to get wiped.

**Drew Rogers** (29:52)

And we're going to look back at these days as, oh,

**Sinclair Toffa** (29:54)

remember when you had to download a separate mobile app for every single Like, that was crazy.

**Drew Rogers** (30:00)

You

**Sinclair Toffa** (30:00)

know, and separate password and separate mobile app and oh, and they didn't... Edit your

**Drew Rogers** (30:05)

home screen.

**Sinclair Toffa** (30:05)

And edit your home screen and... All the things are jiggling around. All the things are jiggling around, yeah. And the mobile apps don't speak to each other. They don't communicate. You have to

**Drew Rogers** (30:15)

face ID into each of them.

**Sinclair Toffa** (30:16)

Yes, to this world where now, you know, you and I and all of us as consumers can, hey, I need this website built, like, find someone in Argentina, great. Pay that person now, right? Spend up a budget, $500, go and pay them. I need to pay for a hotel in Brazil. Boom, coordinate over WhatsApp, booked, done. Businesses are going to get pulled in from consumers, just like the shift with mobile. To adopt agentic payments, right? Agents are now going to make transactions on behalf of people.

And so the businesses need to be prepared to accept payments from agents. And what that comes with is KYA, know your agent, who made this payment, on behalf of which individual or which What is the transaction reporting, for instance? The anti-money laundering? The identity checks that need to be made? What happens when a payment fails, for instance? All of those edge cases, businesses are going to need to account for.

And the existing fintechs that have powered that shift from mobile are not equipped for the agentic wave that's happening right now. Because a lot of those systems were built for humans. And the way we've built Mural has been very much so from a blockchain native standpoint, right? With cryptographic keys, spinning up these highly programmatic wallets, and having that connectivity into the banks. And turns out that that infrastructure is very apt for agents.

Because you can create these highly dynamic wallets with policies on those wallets, and those agents can then spend money from those dedicated accounts. The problem that's happening right now is people are connecting their agents to their plaid, which is probably the worst thing you could do.

**Drew Rogers** (31:58)

Why is that?

**Sinclair Toffa** (31:59)

Because when you connect your agent to your plaid, you have full read-write access to all of your bank accounts. So then the agent can now start

**Drew Rogers** (32:08)

spending. You were just resharing something earlier this week or last week about was it some traditional bank-related thing going on? Was that related to this plaid?

**Sinclair Toffa** (32:17)

So earlier this week, it's been happening over the past couple of weeks, you have all of these banks that are now saying, oh, agents are in customer bank accounts. And I think a lot of the titles

**Drew Rogers** (32:29)

are very

**Sinclair Toffa** (32:29)

clickbaity. But it's a very valid concern on if people are just connecting their banks to these agents, what happens when the agent messes up? Who's responsible? Is it the person? So practically, it's just like, I'm on chat or Grokbot. You're on chat, you're on Instinct, you're on Grokbot, and you say, Yeah, this person. I want to pay this person. You log into Plaid. You make that connection. I'm doing

**Drew Rogers** (32:53)

more and more now where you're making connections. You're like, oh, you don't have my Gmail. You got to have that. But I think maybe what people don't know about Plaid is it's a one-to-many.

**Sinclair Toffa** (33:02)

It's a one-to-many. So Plaid is a one-to-many into all your bank accounts. Which is

**Drew Rogers** (33:06)

why they're so great. Which is why they're so great. For

**Sinclair Toffa** (33:09)

humans. Exactly, for humans. Which is why they're so great for humans. Now, for your agent, you don't want that interaction. You actually want dedicated accounts, dedicated global accounts, where you can just Just allocate money into that global account and then that agent can spend it through a bank transfer, paying someone out via PIX, right? The agent can just pay someone out via PIX in Brazil, SPAY in Mexico, SEPA in Europe, right? Or can do a push-to-card transaction, right?

Or can acquire a transaction via a debit card or credit card, for instance, into the account. And so the agent is kind of this like self-contained fintech. And that's why I say that individuals, people are going to become fintechs as well. Is because they're going to have 3 to 5 to 10 agents each that are carrying out these payments on behalf of them. Very similar to how a business has a finance team carrying out transactions.

Individuals will have their team of agents that are buying things and paying for services on behalf of them.

**Drew Rogers** (34:06)

Personally, like personal concierge. The line, it's going to be really interesting to think about where the line starts and stops between personal versus enterprising. Totally. That's going to be an interesting line to follow.

**Sinclair Toffa** (34:17)

And you have a lot of these new one-man businesses, one-woman businesses that are getting launched nowadays where it's just one person that is, you know, selling things. I mean, you're seeing that line kind of blur now with you have collectibles or social selling companies, right, where you can sell things online and you're basically kind of becoming a business.

**Drew Rogers** (34:38)

Yeah.

**Sinclair Toffa** (34:39)

And you're just selling from your home. Like Whatnot, for instance, right? I

**Drew Rogers** (34:42)

just downloaded Whatnot this week. I was fascinated by it.

**Sinclair Toffa** (34:45)

You know, what I'm hearing is you can sell fish on Whatnot. You can sell meat. You can sell shirt. You can sell pants. So you can become your own store. Like meat from your fridge? Yes. Yes. Well, people are, you know, it's people are selling, people are like, you know, FDA approved fish sellers, for instance, and will catch the fish live. And then we'll be videoing live and then they're selling the fish.

But now what's interesting is like now if you're a whatnot, for instance, then the question is, how do you combat fraud? And how do you handle identity? And how do you handle the compliance? And as you become a more global marketplace, how do you make sure that when people are using agents now to buy things, how do you validate who's real and who's not real? And so that's why we go back to why it's so important for us to be this compliant system of record for our partners is because fraud is going to keep going up.

Cybersecurity is going to keep becoming more important. Trust is going to keep becoming more important. Having direct bank connectivity as we have built it out is going to keep becoming more important for all these reasons.

**Drew Rogers** (35:53)

Speaking of fraud, I've noticed that cards specifically issued cards are becoming like a almost it reminds me of like prepaid cards or gift cards where they can be issued for a specific purpose with a specific budget. And I think companies like Ramp and Brex and like these folks who are like positioned to do that type of business, they seem really positioned well for an agent native transaction economy where they can just like the issue that you brought up of plaid or it's like you're kind of giving everything.

**Sinclair Toffa** (36:23)

Seems contained

**Drew Rogers** (36:24)

within a specific card. I'd love your thoughts on more of this phased approach practically. Who's positioned to win? How is this short-term being accomplished? What's the responsible next step?

**Sinclair Toffa** (36:37)

So the mental model that we use is this concept called fintech centrality. And fintech centrality is how much of a fintech are you? So you have three types, and I'll share how RAM and those types of businesses fit in. You have primary fintechs, secondary fintechs, and tertiary fintechs.

**Drew Rogers** (36:54)

Okay. What's our bodega?

**Sinclair Toffa** (36:56)

Bodega is tertiary. Primary fintech, RAMP would be a primary fintech. A primary fintech is a business that derives most of its revenue from financial services and they are already quite financialized. They already have a lot of financial software.

**Drew Rogers** (37:10)

They're already applying for national trust charters. They're

**Sinclair Toffa** (37:13)

already applying for national trust charters, etc. Those are your primary fintechs. The primary fintechs are most well positioned to move quickly and capture this agentic wave. So when we work with primary fintechs, they move quickly, the integration is quick, and we are an amplifier to their already existing mission, right? And for our RAM, for instance, they are thinking, OK, how do we further build out our blockchain native stack internally, right?

And they're thinking, OK, we're launching these new products worldwide. Let's make sure we have all of that connectivity worldwide, right, as they expand out. They just launched in the UK, for instance. So those are the primary fintechs. The secondary fintechs are ones where they have a portion of their revenue from financial services. An example of a secondary fintech is Deal.

Deal makes most of its money from its EOR services, its HR services, and then a portion of it from the actual financial services of FX, currency conversion, etc.

**Drew Rogers** (38:11)

So like a lot of platforms. A lot of platforms fall into that. Airbnb, Starbucks.

**Sinclair Toffa** (38:16)

Airbnb. Exactly. Yep. Airbnb falls into that. Some of the food delivery companies would fall into that. Maybe Uber as well might fall into that. We are spending a lot of time on the primaries and a majority on the secondary fintechs because the secondary fintechs are the ones who Already are earning some money but really need to amplify that and building great financial infrastructure and software is not their expertise.

And they view it as a way to launch new products or expand to new geographies and acquire more customers or retain more customers. Back to our initial point of When you're generating the value, you want to keep the value. You want to own the account. Then you have the tertiary fintechs, which are really interesting. The tertiary fintechs are ones where you're like, no way this would become a fintech. And these are the ones, they don't have financial software. They're not really making money from financial services.

So an example is, you know, the bodega, this company that's connecting all the bodegas, for instance, that's an example of a tertiary fintech.

**Drew Rogers** (39:15)

Another example

**Sinclair Toffa** (39:16)

of a tertiary fintech could be, for instance, Spotify. It could be an example. How do you

**Drew Rogers** (39:21)

differentiate Spotify from secondary?

**Sinclair Toffa** (39:24)

Secondary is they're making a portion of their revenue from financial services. They're

**Drew Rogers** (39:29)

already categorized in that way?

**Sinclair Toffa** (39:31)

Exactly. They're already categorized that way. They might have some form of an account or a wallet. They're making some revenue for an exchange, for instance. People

**Drew Rogers** (39:39)

have Spotify accounts, but we don't have wallets yet.

**Sinclair Toffa** (39:41)

Yeah, for Spotify, it's you as a consumer, you're listening to Spotify, right? But what about the artists, for instance, where one day you can imagine, you know, Spotify wallet, where as an artist, you're getting paid in real time, or as a podcaster, you're getting paid in real time into your Spotify wallet. And you can keep the money in there, right? You can retain it in there and you can grow it over time and And then you have other, you know, other businesses in like supply chain, for instance, right?

Supply chain businesses, transportation businesses, which are tertiary fintechs. They're moving physical goods. They are moving a lot of money, but they're not, you know, they're not retaining it in the ecosystem. And so those tertiary fintechs are in the process of becoming secondaries and then into primaries.

**Drew Rogers** (40:25)

So it sounds like primaries, in your view, have the right to move really quickly. Yes. They're positioned best for this like agent-native economy. What would be your...

**Sinclair Toffa** (40:35)

I will caveat with, you know, you have the ramps, the Brexits of the world, etc. Seems so competitive though. It's very competitive. It's a knife fight out there. This seems

**Drew Rogers** (40:45)

different, though. This seems more like green.

**Sinclair Toffa** (40:48)

You've got these AI-native tertiary fintechs, which are basically companies that are fully AI-native that are going into these very niche fields on supply chain.

**Drew Rogers** (41:02)

It's like the general catalyst kind of

**Sinclair Toffa** (41:03)

approach. Exactly. And the ones who are fully AI native are actually really well positioned to integrate these new financialized accounts to rapidly build products. And so we're going to see class of these new AI native businesses, their growth is going to outpace a lot of these existing primary fintechs.

**Drew Rogers** (41:23)

What would be your advice to both sides? You're meeting with someone high up at a ramp or Mercury or Brex or whatever. What is the strategy to win the knife fight and win? And then what would be your cautions there? And then also for the tertiary, the opportunity that you would give to that side of the spectrum?

**Sinclair Toffa** (41:43)

Yeah, for the primaries, I believe oftentimes the primary fintechs underestimate the complexity of connecting into blockchain-based systems, for instance. And they underestimate the complexity of creating wallets, managing custody, doing all the transaction monitoring, the anti-money laundering, and connecting all of these newer real-time blockchain-based systems to the existing fiat-based systems, global fiat systems. So they're

**Drew Rogers** (42:07)

trying to build a lot of this themselves.

**Sinclair Toffa** (42:08)

It's always a build versus buy question.

**Drew Rogers** (42:11)

You're seeing a lot of them try to build, think they can, but actually maybe they should buy.

**Sinclair Toffa** (42:16)

Correct. I'm seeing a lot of them try to build, build a lot of the things in-house, realize it's a lot more complex than they assumed, and then the go-live launch keeps getting pushed back and over and over again. Where does

**Drew Rogers** (42:28)

complexity find itself over there?

**Sinclair Toffa** (42:30)

The way they might try to approach it is they might try to stitch together different providers, stitch together different platforms, right? And then by stitching together all these different platforms, the complexity actually keeps going up. And you get to a point where you realize, oh, wait, payments are being made over this blockchain network of Ethereum, but it's not compatible with Tempo, which is not compatible with Polygon.

Oh we're trying to do these named accounts in Brazil or in Colombia but the named accounts are only limited to individuals and you can't set up business named accounts and the account number is done this way and oh wait there's this additional local tax that applies for these types of digital assets stablecoin payments we didn't realize that so all of these nuances start piling up oh the local regulation is changing and so then it keeps piling up and it delays the launch.

And so my advice for these ones is in the build versus by conversations, go through the process of actually architecting it on a whiteboard. And this is what we love to do is like, let's look at your existing stack, right? Let's look at where you are now, where you want to be.

**Drew Rogers** (43:33)

Now we're back to the ontology mindset.

**Sinclair Toffa** (43:35)

We're back to the ontology mindset, right? Where you're at, where do you want to be? Let's talk about your core ontology. You want to launch these new products within a month or by end of year. Great. Here is where we would insert the Mural platform and how it would connect. And here's the path of building with us. And then here's the path of just trying to build it internally. And this is the same discussion we always had at Palantir.

**Drew Rogers** (43:58)

Are you doing this in person? Like you're going to the office? We're

**Sinclair Toffa** (44:00)

doing this in person. We're showing up in person with a couple of our four deployed engineers. We're mapping it

**Drew Rogers** (44:05)

out. Like a day-long meeting?

**Sinclair Toffa** (44:07)

It's usually half a day or a day-long meeting. Wow. So we have, you know, a couple of FTE teams deployed right now in Europe. We have some in Colombia right now as well, in Argentina, right? They're constantly traveling around. And that's what I would spend more time upfront doing.

**Drew Rogers** (44:21)

Yeah. Instead of like, let's just go launch the pilot we build ourselves in this country for this use case for this chain. Correct, because

**Sinclair Toffa** (44:28)

the complexity just keeps going up. And then on the other side, for the tertiary fintechs, for the AI-native businesses that are launching, the new AI-native browsers, for instance, et cetera, my advice for them is what's most important is growth, acquisition, and retention as you scale up your business. And being able to partner with a financial software business like us allows you to focus on the distribution, focus on the experience, and capture the market very quickly.

And the market's turning very quickly now as we speak. And so speed is a competitive advantage. And having that system of record is really important so that you can quickly launch new products and avoid the fragmentation and the stitching together. And it's going back to the architecture topic, which is for you as a startup, what is your core ontology? What are your core business objects? What do you want to own?

And how do you make sure that the way you build it is in a way Drew Rogers, Sinclair Toffa, Zach Fowler, Stabledash Global, Stabledash Live. Drew Rogers, Sinclair Toffa, Zach Fowler, Stabledash

**Sinclair Toffa** (45:40)

At the same time, trust and cybersecurity is going to become more and more important, especially at the intersection of fintech and agentic. We feel very blessed to be building and leading at the forefront, and we feel like now is a really amazing opportunity for those who are building AI-native businesses to become fintechs now sooner than later, and for the existing ones to speed up their go-to-market by partnering.

**Drew Rogers** (46:05)

Yeah, so cool. Sinclair, thanks for the time. I learned a lot during this conversation, so thanks for sharing.

**Sinclair Toffa** (46:10)

Thank you. I enjoyed

**Drew Rogers** (46:11)

it. We're going to go take a walk outside. I think during our dinner that we had last time, I spilled wine all over Sinclair's pants. It was like

**Sinclair Toffa** (46:22)

the first two minutes.

**Drew Rogers** (46:23)

Yeah, it was like the first two minutes. I had

**Sinclair Toffa** (46:24)

bought those pants the day before. You no way.

**Drew Rogers** (46:26)

So we're going to go take a walk, and I'm going to buy you a new pair of pants. Wait, wait. There's a kids store across the street. No,

**Sinclair Toffa** (46:31)

that's too nice. We're going to end here. We're going to go

**Drew Rogers** (46:33)

get

**Sinclair Toffa** (46:33)

some

**Drew Rogers** (46:34)

pants. All right, that's what we're doing.
