TRANSKRİPT 🔗
WEBVTT
00:00:00.320 --> 00:00:08.320
Passage of the stablecoin legislation drafted by the Senate, dubbed the Genius Act, because analysts say a wave of competition can complicate things.
00:00:13.599 --> 00:00:16.160
Welcome back everyone to the Stable Pulse.
00:00:16.160 --> 00:00:18.079
My name is Dante Reminick.
00:00:18.079 --> 00:00:23.839
We have a very special guest here and a really, really well-timed episode, I feel.
00:00:23.839 --> 00:00:26.399
So without further ado, I'm just going to get right into it.
00:00:26.399 --> 00:00:28.239
Dileep, welcome to the show.
00:00:28.239 --> 00:00:34.240
As you know from being a longtime listener, we always start with an episode or a segment that we call Braggadocious.
00:00:34.240 --> 00:00:39.759
We use this as an opportunity for you to brag about yourself and all of the really awesome stuff that you have built.
00:00:39.759 --> 00:00:43.280
You have a lot of reason to brag, uh, especially recently.
00:00:43.280 --> 00:00:45.119
So I'm going to sit back and shut up.
00:00:45.119 --> 00:00:46.159
Tell us about yourself.
00:00:46.159 --> 00:00:47.600
Tell us about what you've built.
00:00:47.600 --> 00:00:50.320
Please, please, please do not be humble.
00:00:51.039 --> 00:00:53.119
Well, forest, thanks for having me on, Dante.
00:00:53.119 --> 00:00:56.320
I've heard a lot uh about the show, and so I'm really excited to be here.
00:00:56.320 --> 00:01:00.240
Um I'm Daliep, I'm the founder and CEO of a company called Jeeves.
00:01:00.240 --> 00:01:06.799
Um, this round that we are just announcing, which was about two days ago, uh, was a $110 million round.
00:01:06.799 --> 00:01:14.719
And what's really interesting is that we have always been building in the financial infrastructure ecosystem focused on enterprises.
00:01:14.719 --> 00:01:17.439
But this round is really for us to go deeper into stablecoin.
00:01:17.439 --> 00:01:19.519
So taking a step back, what do we do?
00:01:19.519 --> 00:01:24.000
So Jeeves provides a financial operating system for global enterprises.
00:01:24.000 --> 00:01:25.120
And what does that mean?
00:01:25.120 --> 00:01:37.439
If you're a company like BMW or uh Lululemon or H ⁇ M that uses Jeeves today, what we give you is a uh combined operating system that includes cards in 25 to 30 countries.
00:01:37.439 --> 00:01:43.040
So you do not need a corporate card for your employees in the US, plus another one in UK, another one in Brazil, etc.
00:01:43.040 --> 00:01:44.079
All of that we do.
00:01:44.079 --> 00:01:48.560
And then we also do payments in and out to 190 countries, and we just launched accounts receivable.
00:01:48.560 --> 00:01:59.359
And so really the pitch is if you're a company that operates in more than one country, we can combine your stack and fix it and make it much, much easier so you get kind of one login and everything kind of lives together.
00:01:59.359 --> 00:02:02.079
And we do own a lot of infrastructure to be able to do that.
00:02:02.079 -->
Load full transcript
00:02:10.080
Now, on to the side about things to brag on, um, I would say for us, it's really been impressive to see the growth in the last eight months.
00:02:10.080 --> 00:02:19.280
We really launched our stablecoin product in January, and we just crossed a billion and a half in annualized TPV, so that's north of 130, 140 million a month in eight months.
00:02:19.280 --> 00:02:25.280
And I think you know, a lot of times when you think about stablecoin use cases, they tend to be focused on uh small businesses.
00:02:25.280 --> 00:02:30.560
So Pfypers and e-commerce company, the digital nomad, uh sometimes it's PSPs and FinTechs.
00:02:30.560 --> 00:02:31.680
We sell to enterprises.
00:02:31.680 --> 00:02:36.960
So it's the actual opposite of that, where you've got to go sell the full enterprise sales cycle, and then stablecoin is part of that.
00:02:36.960 --> 00:02:41.039
So for me, that's something that we're very proud about and you know I think is really important.
00:02:41.039 --> 00:02:48.879
And then I think the other side that is also very important is we're one of the few companies that combine both fiat and stablecoin.
00:02:48.879 --> 00:02:53.520
We have real infrastructure, we have real licenses, we have three principal memberships for card issuing.
00:02:53.520 --> 00:02:57.599
Uh we can do payoffs to 190 countries, and on top of that, we build the stablecoin layer.
00:02:57.599 --> 00:03:05.439
And to be able to do that, so far, we've moved about 5 billion in annualized TPV, and that's grown about 10x in the last two years.
00:03:05.439 --> 00:03:09.840
So the business is clearly hitting this inflection point, which is why uh we did this round.
00:03:09.840 --> 00:03:12.800
Uh me personally, I'm a second-time founder, so this is my second company.
00:03:12.800 --> 00:03:23.199
Before this, I started a company in marketing automation space, hated the space, but sold the company um for about 100 million plus, and we raised about 19 million, very different time, but good outcome for everybody.
00:03:23.199 --> 00:03:32.960
Uh and then the problem I'm trying to solve with Jeeves was something I'd seen in that company, which is if you operate in two countries, we did, you needed two different bank accounts, two different corporate cards, two different payment reels.
00:03:32.960 --> 00:03:38.080
And then at the end of the month, when I say, hey, you know, CFO or my team, like how much did we spend last month?
00:03:38.080 --> 00:03:38.719
Nobody knows.
00:03:38.719 --> 00:03:40.639
Takes you 30 days to answer that question.
00:03:40.639 --> 00:03:47.759
And so that was this core problem pre-stable coin that was just kind of like haunting me, but I'm like, I have to figure out how to solve this.
00:03:47.759 --> 00:03:51.199
Uh and that's the genesis of what kind of Jeeves is today.
00:03:52.000 --> 00:03:52.479
I love that.
00:03:52.479 --> 00:03:54.960
First of all, that's a fantastic introduction.
00:03:54.960 --> 00:03:57.120
I have
many, many questions.
00:03:57.120 --> 00:04:01.520
And so I hope that we can I hope that we can keep this efficient, but I doubt that we'll be able to.
00:04:01.520 --> 00:04:02.080
Yeah, no.
00:04:02.080 --> 00:04:09.599
First, I mean, you guys sell to major, major enterprises that have absolutely nothing to do with stable coins, right?
00:04:09.599 --> 00:04:11.599
You mentioned Lululemon, you mentioned BMW.
00:04:12.000 --> 00:04:13.280
They don't care enough about stable coins.
00:04:13.520 --> 00:04:15.599
They they don't care about stable coins, right?
00:04:15.599 --> 00:04:20.480
When you are going to pitch these folks, do you talk about stable coins?
00:04:20.480 --> 00:04:26.319
Or are you just pitching better financial services that just so happen to be powered by stable coins in the bank?
00:04:26.560 --> 00:04:28.000
It's it's it's exactly the latter.
00:04:28.000 --> 00:04:32.480
And and if you think about it, when you go pitch we do payments, nobody says we do ACH.
00:04:32.480 --> 00:04:34.800
You just say we do payments, and like the payments show up.
00:04:34.800 --> 00:04:38.720
No one's like, okay, what's the protocol and the rail that you it's just one, do I trust you?
00:04:38.720 --> 00:04:39.839
Two, what's the cost?
00:04:39.839 --> 00:04:44.399
Three, is the money gonna show up or am I gonna get fired as a CFO because the money never showed up?
00:04:44.399 --> 00:04:45.759
Like that's what they care about, right?
00:04:45.759 --> 00:04:46.879
And that's what we sell.
00:04:46.879 --> 00:04:50.319
And and it's interesting because we don't hide stablecoin, it's just abstracted away.
00:04:50.319 --> 00:04:58.079
And so what that means is one of the things that we do, which is probably the hardest thing we do, is we maintain our own ledger in 25 countries.
00:04:58.079 --> 00:05:05.279
So that means at the end of the month, when you say, hey, the leap, give me a statement for my account, that statement line by line is generated by Gs.
00:05:05.279 --> 00:05:07.519
We don't go to a third party, get it, and send it to you.
00:05:07.519 --> 00:05:14.959
And it sounds easy, but that's actually the hardest thing we do, because that means your ledger, your data, your FX, this is 25 countries.
00:05:14.959 --> 00:05:16.800
So multiple currencies has to be perfect.
00:05:16.800 --> 00:05:22.480
And if it isn't, you're gonna hear that in the first like 30 minutes because someone's gonna tell you the numbers off it and I'm not gonna pay that.
00:05:22.480 --> 00:05:22.720
Right.
00:05:22.720 --> 00:05:25.120
And so that that's really the muscle we built.
00:05:25.120 --> 00:05:29.040
And because we built that ledger, adding stablecoin is just another currency.
00:05:29.040 --> 00:05:32.160
And so we were able to make the ledger just 10 times more efficient.
00:05:32.160 --> 00:05:35.279
And to your point, we don't sell, hey, come to us for stablecoin.
00:05:35.279 --> 00:05:37.839
We say we actually call the product Jeeves Instant
Pay.
00:05:37.839 --> 00:05:43.759
Uh so it's stablecoin, but like the product that we sell is Jeeves Instant Pay, and we say, hey, you can use regular or you can use instant pay.
00:05:43.759 --> 00:05:45.120
Here's a cost, here's a price, blah, blah.
00:05:45.120 --> 00:05:46.000
And then you pick.
00:05:46.000 --> 00:05:49.600
And as you can probably expect, you know, people tend to pick instant pay.
00:05:50.240 --> 00:05:50.720
Interesting.
00:05:50.720 --> 00:05:58.639
And so a lot of your competitors that are competing with you for deals like BMW and Lululemon, are they also using stable coins?
00:05:58.639 --> 00:06:05.839
Or are you guys able to go in there and offer better financial technology because you are using stable coins when your competitors are not?
00:06:06.240 --> 00:06:09.839
Yeah, so it's it's very interesting because if you like if you take stablecoin payments, right?
00:06:09.839 --> 00:06:16.000
And and this is kind of the the the cry that everyone's hearing right now, which is FX fees are getting compressed, right?
00:06:16.000 --> 00:06:21.600
Like if you think about the off-ramp from USD or USDC to USD, that used to be 1%.
00:06:21.600 --> 00:06:22.879
Now it's five bips.
00:06:22.879 --> 00:06:25.439
It's gonna be zero, like complete compression, right?
00:06:25.439 --> 00:06:26.720
That's a hundred bips that you've lost.
00:06:26.720 --> 00:06:28.079
FX, same thing's happening.
00:06:28.079 --> 00:06:29.600
You'll you're gonna lose all your FX.
00:06:29.600 --> 00:06:33.279
It's gonna be some fees, but it's nothing like the fat spread that that's there.
00:06:33.279 --> 00:06:39.040
It's actually great for us because what we sell is not the FX, it's the entire workflow automation before the FX.
00:06:39.040 --> 00:06:49.680
So, as an example, in Mexico, we have a product where you can put in your tax ID, we'll go download all your receipt, uh, your invoices from the central database, fill in the information, you hit a button, we'll send the payment.
00:06:49.680 --> 00:06:53.199
The last step, that's the payment, and that's FX and all of that stuff.
00:06:53.199 --> 00:06:56.240
But we charge you for everything else before the last step.
00:06:56.240 --> 00:07:02.319
So we can actually subsidize the cost of FX, you know, give it away close to free, because that's not where we make money.
00:07:02.319 --> 00:07:04.639
We make money on all the other steps before that.
00:07:04.639 --> 00:07:11.680
And that's the difference between the product that we have, even pre-stablecoin, which is we sell financial automation, we sell efficiency gains.
00:07:11.680 --> 00:07:14.800
That's what a CFO of a 5,000 person company cares about.
00:07:14.800 --> 00:07:19.920
They don't care about five bips or 10 bips here or there, which is just a very different audience from like an SB.
00:07:21.040 --> 00:07:21.600
Interesting.
00:07:21.600 --> 00:07:29.519
And so are you guys able to do FX better and compress those rates more because
you're built on a stablecoin stack?
00:07:29.920 --> 00:07:33.360
Yeah, so we do the classic kind of stablecoin sandwich, so to speak.
00:07:33.360 --> 00:07:34.399
So I'll give you an example.
00:07:34.399 --> 00:07:40.879
Previously, two years ago, you could have come to Jeeves and say, hey, I want to do a payment between uh, let's say Mexico and Europe, right?
00:07:40.879 --> 00:07:48.800
And you'll say, great, put in your information for CEPA, hit send, we collect your funds in Mexico, send it to correspondent bank one, send the second bank, and then they pay it out.
00:07:48.800 --> 00:07:50.319
Takes about one to two days.
00:07:50.319 --> 00:07:53.279
Now you come to Jeeves, it actually looks exactly the same.
00:07:53.279 --> 00:07:55.680
It's the same information, it's still SEPA.
00:07:55.680 --> 00:07:58.959
You hit send, but now what happens is we collect your funds in Mexico.
00:07:58.959 --> 00:08:03.680
As soon as we do, we release USDC or USDT, whatever you want, and then we pay out on the other side.
00:08:03.680 --> 00:08:04.879
It takes one to two hours.
00:08:04.879 --> 00:08:07.199
But everything else looks exactly the same.
00:08:07.199 --> 00:08:10.639
Same platform, same statements, blended with your cards.
00:08:10.639 --> 00:08:11.439
That's what we sell.
00:08:11.439 --> 00:08:17.759
It just so happens, exactly like you're saying, that the stack that's you know providing the money movement is stablecoin.
00:08:17.759 --> 00:08:20.879
But we don't go in and say, hey, come to us just to access the stablecoin.
00:08:20.879 --> 00:08:22.079
That's that's not where we're even.
00:08:22.079 --> 00:08:26.560
There are many companies that are very good at that, but that's not like where I see our differentiation.
00:08:27.920 --> 00:08:33.919
Is it fair to say that you're fighting to be the primary financial relationship with these large enterprises?
00:08:34.320 --> 00:08:39.360
I we want to be the primary financial relationship slash operating system.
00:08:39.360 --> 00:08:45.600
It's it's actually a very good point because the number one uh competition that we see is usually a local bank.
00:08:45.600 --> 00:08:50.480
So these companies they tend to have a private bank or they tend to have someone they can call and say, hey, I need to send this money.
00:08:50.480 --> 00:08:52.879
And the guy's like, great, like let me help you with that, blah, blah, blah.
00:08:52.879 --> 00:08:53.919
Because these are bigger companies.
00:08:53.919 --> 00:08:54.960
And we say, keep that.
00:08:54.960 --> 00:08:55.840
Like, don't lose that.
00:08:55.840 --> 00:08:56.960
That's amazing, keep it.
00:08:56.960 --> 00:09:04.559
But what we can give you that your bank is not going to give you is unlimited virtual cards, which in a country like Brazil is actually very important because there's a lot of fraud.
00:09:04.559 --> 00:09:06.879
And so we have this product that's single-use cards.
00:09:06.879 --> 00:09:09.440
So companies can just literally call the card.
00:09:09.440 --> 00:09:12.480
You don't need a human in the loop, it's agentic, API, whatever you want.
00:09:12.480 --> 00:09:16.240
And the card just spins up, gives you a number, you use the number, and the card dies.
00:09:16.240 --> 00:09:21.039
And so there's very minimal fraud because you can't use the card, and the reconciliation is very high.
00:09:21.039 --> 00:09:23.519
And we have companies that will do 100,000 cards a month.
00:09:23.519 --> 00:09:25.279
A bank is never gonna give you that, right?
00:09:25.279 --> 00:09:30.720
And so that's how we compete, which is like keep the bank, we're the operating layer, we're the layer that sits on top of all the banks.
00:09:30.720 --> 00:09:31.759
And you should keep your bank.
00:09:31.759 --> 00:09:32.320
Banks are great.
00:09:32.320 --> 00:09:33.759
We work with them, we have a lot of bank camps.
00:09:33.759 --> 00:09:34.480
That's awesome, right?
00:09:34.480 --> 00:09:42.799
But like what they're not gonna do is say six things, they're not gonna give you cards in Portugal for Brazilian business, the Spanish cards for Mexican business, because that's not where they win.
00:09:42.799 --> 00:09:43.759
That's where we went, right?
00:09:43.759 --> 00:09:48.000
And so it again, this these the spaces we compete in, it's it's so big.
00:09:48.000 --> 00:09:50.240
It's not a winner-take-all market.
00:09:50.240 --> 00:09:52.399
It's not like it's you know one versus the other.
00:09:52.399 --> 00:09:54.159
These are trillion dollar spaces, right?
00:09:54.159 --> 00:09:56.720
And you win one, two, three percent.
00:09:56.720 --> 00:09:59.919
I mean, you're talking about a massive, massive, massive company.
00:10:01.200 --> 00:10:02.639
So you don't make your money on FX.
00:10:02.639 --> 00:10:04.240
Where do you make your money?
00:10:04.559 --> 00:10:06.799
We we we actually it's really interesting.
00:10:06.799 --> 00:10:09.519
Um if you look at our revenue stream, right?
00:10:09.519 --> 00:10:11.600
And and and just for classification, right?
00:10:11.600 --> 00:10:15.360
We're we're falling firmly into like that expense management box, right?
00:10:15.360 --> 00:10:20.320
Like that's that's our core product is cards, we have payments, just the standard things, and we move it around, blah, blah.
00:10:20.320 --> 00:10:25.440
What's really interesting is for us, interchange is less than 40% of the business.
00:10:25.440 --> 00:10:29.600
Most expense management companies, interchange is 80 to 90% of your business.
00:10:29.600 --> 00:10:33.679
That that's the that's the biggest chunk of your revenues, interchange BRECS, et cetera, et cetera.
00:10:33.679 --> 00:10:35.120
For us, it's less than 40%.
00:10:35.120 --> 00:10:41.679
And so it's a very, very blended revenue stream where 40% is cards, so that's interchange plus our cards API.
00:10:41.679 --> 00:10:44.320
40% is payments, but payments is not FX.
00:10:44.320 --> 00:10:46.399
It's the workflow automation
that we charge for.
00:10:46.399 --> 00:10:49.519
It includes FX, but like that's not the big driver, so that's about 80%.
00:10:49.519 --> 00:10:54.720
And then we charge a lot of fees, um, about 15 to 18% is fees.
00:10:54.720 --> 00:11:07.840
So it looks a little bit more like a revolute than uh what I call it is like an enterprise version of Revolute where you have this massive engine that's very profitable, that's growing, and then you get to take all these other bets and then subsidize it and kind of eat the market share.
00:11:07.840 --> 00:11:13.360
Is that your margin is my, you know, like whatever the the thing was from your margin is my opportunity.
00:11:13.360 --> 00:11:14.240
My opportunity.
00:11:14.240 --> 00:11:15.600
Because I can subsidize it, right?
00:11:15.600 --> 00:11:17.039
Because I'm making money somewhere else.
00:11:17.120 --> 00:11:23.840
So maybe I I'm approaching this the wrong way and trying to categorize you when I should not be.
00:11:23.840 --> 00:11:30.639
But like I think that when I hear people talking about Jeeves, oftentimes they use the word neobank, right?
00:11:30.639 --> 00:11:32.799
And so to your point, you mentioned Revolute.
00:11:32.799 --> 00:11:38.240
You just mentioned that you're an expense management platform, much more akin to something like ramp.
00:11:38.240 --> 00:11:46.159
But in reality, you're this like weird, sexy hybrid between between the two, right?
00:11:47.200 --> 00:11:53.200
You'll probably figure that out faster than a lot of investors have dead of it, which is it's it's it's it's exactly that.
00:11:53.200 --> 00:11:55.279
Like it's everyone's like, okay, what box is it?
00:11:55.279 --> 00:11:58.000
But the thing is now the boxes have changed.
00:11:58.000 --> 00:11:59.600
I mean, like Ramp is an AI company.
00:11:59.600 --> 00:12:05.519
Like, like the boxes, there's no one specific, you know, thing of like, oh, I get it, you're exactly this we're not.
00:12:05.519 --> 00:12:06.879
Like we operate in 35 countries.
00:12:06.879 --> 00:12:07.840
We're 200 people.
00:12:07.840 --> 00:12:09.360
Like, not a lot of companies do that.
00:12:09.360 --> 00:12:11.039
We do card issuing directly.
00:12:11.039 --> 00:12:12.399
I mean, we are the issuing entity.
00:12:12.399 --> 00:12:15.039
It's like you get a card that says issued by Jeeves, not a bank.
00:12:15.039 --> 00:12:15.759
That's us.
00:12:15.759 --> 00:12:17.200
Like we issue the cards directly.
00:12:17.200 --> 00:12:21.919
Like we go very, very deep in the stack, and yes, part of it is expense management, part of it is payments.
00:12:21.919 --> 00:12:25.919
But it's very hard to just put into this one box and be like, what is this company?
00:12:25.919 --> 00:12:32.000
And I think most successful companies in the next five years, the ones you see, the ones that are growing very fast, they're gonna look like this.
00:12:32.000 --> 00:12:37.919
They're gonna pick the best parts of what worked for them and have this hybrid product because that's what's also now changed with AI.
00:12:37.919 --> 00:12:40.960
You just can build so much more than you could before.
00:12:40.960 --> 00:12:42.000
It's actually very interesting.
00:12:42.000 --> 00:12:46.480
I had this conversation internally, uh, and we were using this metaphor of like at bats.
00:12:46.480 --> 00:12:48.480
If you think about it three years ago, right?
00:12:48.480 --> 00:12:51.759
And and like let's say the euro is starting, you're like, hey, we're gonna do two bets.
00:12:51.759 --> 00:12:53.519
We're gonna do A and we're gonna do B.
00:12:53.519 --> 00:12:55.360
And we're gonna spend all this time, we're gonna build it out.
00:12:55.360 --> 00:12:58.879
I need at least one of them to work just to make the Euro worthwhile, right?
00:12:58.879 --> 00:13:00.159
So I need one out of two.
00:13:00.159 --> 00:13:02.240
Now with AI, I can do 10 bets.
00:13:02.240 --> 00:13:03.759
And I I if I get three of those.
00:13:04.240 --> 00:13:05.440
You still only need one or two to work.
00:13:05.440 --> 00:13:06.399
Exactly, right?
00:13:06.480 --> 00:13:09.360
So like the amount you can do is so much more.
00:13:09.360 --> 00:13:13.279
And if I get three of that, that means my output is three better than it was three years ago.
00:13:13.279 --> 00:13:14.399
Like that's what's changed.
00:13:14.399 --> 00:13:14.720
Yeah.
00:13:15.039 --> 00:13:19.840
And so are you guys just constantly running experiments internally and trying to poke and prod?
00:13:19.840 --> 00:13:21.919
Like, tell me about that experimentation internally.
00:13:22.240 --> 00:13:26.240
So our core philosophy is that we want to own the CFO suite.
00:13:26.240 --> 00:13:31.919
So what that means is you have carts, which is our core wedge, uh, then you have AP, which is payments.
00:13:31.919 --> 00:13:34.240
80% of companies use more than one product line.
00:13:34.240 --> 00:13:34.879
So that's two.
00:13:34.879 --> 00:13:37.840
We just launched AR, which is invoicing, so accounts are receivable.
00:13:37.840 --> 00:13:40.399
And and this is kind of that you know virtuous circle, right?
00:13:40.399 --> 00:13:45.120
If you use us for invoicing and the funds come to us, then we have your funds because if we have your funds, we can give you more spend.
00:13:45.120 --> 00:13:46.799
Because if you get more spend, you can do more payments.
00:13:46.799 --> 00:13:48.559
So you get like it's that whole cycle.
00:13:48.559 --> 00:13:53.200
So you have cards, you have AP, you have AR, then we want to add treasury, and we want to add payroll.
00:13:53.200 --> 00:13:55.519
And then we have this concept of what we're releasing.
00:13:55.519 --> 00:13:59.440
It's still something we're working on, which is kind of this Gee's AI services module.
00:13:59.440 --> 00:14:05.679
But the idea is that we will be your entire CFO suite, back office, all in one in any country that you operate.
00:14:05.679 --> 00:14:06.399
That's the goal.
00:14:06.399 -->
00:14:08.480
Like, and then it's a big goal, right?
00:14:08.480 --> 00:14:11.120
Because you're gonna do this in like 30, 40, 50 countries.
00:14:11.120 --> 00:14:12.000
But that's the goal.
00:14:12.480 --> 00:14:15.840
Well, I to that point, I think we're getting ahead of ourselves a little bit.
00:14:15.840 --> 00:14:24.879
Like, we're talking about experimenting and building out this full CFO suite filled with payments and multiple different jurisdictions, and you're acting like it's relatively easy.
00:14:24.879 --> 00:14:31.039
You're like, yeah, with AI, we can just, you know, we can go and build it, but you can't really vibe code what you're what you're talking about.
00:14:31.039 --> 00:14:34.159
Can you walk me through what it takes to build it?
00:14:34.159 --> 00:14:39.919
Like, frankly, I just want to be very, very clear with our audience that you make it sound easy, but it's not.
00:14:39.919 --> 00:14:41.759
It's very, very difficult.
00:14:42.480 --> 00:14:43.440
It is not easy.
00:14:43.440 --> 00:14:46.240
This is it, it's what it's it's really, really fucking hard.
00:14:46.240 --> 00:14:50.480
Like it's it's uh, you know, like you have to go get the licenses, you have to build a stack.
00:14:50.480 --> 00:14:55.120
So just to give you an idea, and I'll use card issue, because I think that's the clearest one to kind of explain, right?
00:14:55.120 --> 00:14:59.600
So just to give the background, I don't have a fintech background, I don't have a banking background.
00:14:59.600 --> 00:15:04.879
And I think part of the reason Jeeves is built the way it is, is because I didn't see it the same way, good or bad.
00:15:04.879 --> 00:15:14.799
Candidly, we probably bit off much more than we could chew in the first go-around of this, but because we survived and because we made it to this point, it's a very defensible mode.
00:15:14.799 --> 00:15:19.679
Because if someone else wants to do it, they have to go to 25 countries, get the licenses, get the stack.
00:15:19.679 --> 00:15:22.639
But it's really because I the way we had looked at it was a little bit different.
00:15:22.639 --> 00:15:23.440
So, what does that mean?
00:15:23.440 --> 00:15:25.360
So if you look at like card issuing, right?
00:15:25.360 --> 00:15:27.200
So that's our core product that we start with.
00:15:27.200 --> 00:15:28.639
It consists of four components.
00:15:28.639 --> 00:15:42.080
You have the top of the funnel, KYB, customer acquisition, you have the processor, you have the program manager, which is a regulation compliance layer, and then you have the issuing bin, the bank identification number, the 16 digits Visa MasterCard, give that to you, and that's how you issue the card locally.
00:15:42.080 --> 00:15:47.039
What we did that's different from day one is the first three sit on our own stack.
00:15:47.039 --> 00:15:50.240
So it works across partners, et cetera, but it sits with us.
00:15:50.240 --> 00:15:55.440
So what that means is when we went from Mexico to Colombia, the only thing we needed to change was that bin.
00:15:55.440 --> 00:15:58.799
When we go from Colombia to Brazil, the only thing we need to change was the bin.
00:15:58.799 --> 00:16:02.799
When we go to Indonesia, which we don't cover today, the only thing we need to change is that bin.
00:16:02.799 --> 00:16:08.960
That's structurally different from how most fintechs would build it, because most fintechs are only focused on one country.
00:16:08.960 --> 00:16:12.000
If you're only doing one country, why the F would you build this entire stack?
00:16:12.000 --> 00:16:12.720
It makes no sense.
00:16:12.720 --> 00:16:15.200
It's an immense amount of overhead for one country.
00:16:15.200 --> 00:16:17.360
And so usually you go very deep in one country.
00:16:17.360 --> 00:16:20.639
Three years later, you're like, let's do the next one, build it again, next one, build.
00:16:20.639 --> 00:16:24.480
So ours is just fundamentally different, good or bad, but that's the core difference.
00:16:24.480 --> 00:16:33.360
Like we built this very modular from day one because our goal, our thesis was that companies are going global from day one, but banking is still country and currency specific.
00:16:33.360 --> 00:16:34.159
How do you solve it?
00:16:34.159 --> 00:16:37.279
The core issues, the data is the data is in different silos.
00:16:37.279 --> 00:16:40.320
So if you want to get the data, you have to build this abstraction layer on top.
00:16:40.320 --> 00:16:41.519
That's our core IP.
00:16:41.519 --> 00:16:43.679
So that that's that's the difference.
00:16:43.679 --> 00:16:48.000
I don't know if that's helpful, but like that that's really what's the difference between how we're built.
00:16:48.399 --> 00:16:49.440
I think that's super helpful.
00:16:49.440 --> 00:16:52.240
I think you explained why it's so hard to build.
00:16:52.240 --> 00:16:54.000
Like there's sort of two different layers here.
00:16:54.000 --> 00:16:58.960
There's like, okay, the the financial primitives, who's my sponsor bank?
00:16:58.960 --> 00:17:01.759
What am I doing with bin numbers, et cetera, et cetera?
00:17:01.759 --> 00:17:03.039
But that's step one.
00:17:03.039 --> 00:17:05.119
Like it's a very hard step one, but it's step one.
00:17:05.119 --> 00:17:14.319
And then step two is okay, how do I create the data layer on top of it to properly serve everything that a CFO needs?
00:17:14.319 --> 00:17:16.640
AP, AR, all of that stuff.
00:17:16.640 --> 00:17:18.079
Is that is that fair?
00:17:18.559 --> 00:17:19.119
That's fair.
00:17:19.119 --> 00:17:24.640
That layer is immensely um it it it's that there's a lot of trial and error there.
00:17:24.640 --> 00:17:29.519
I that there is and and this is why it's always funny when people are like, oh, you can make a 10-minute neo bank.
00:17:29.519 --> 00:17:31.440
The hard part is not spitting up the neobank.
00:17:31.440 --> 00:17:34.559
The hard part is the operational side of running a new bank.
00:17:34.559 --> 00:17:38.319
That's that's what like doing payments in 190 countries, that's the skill set we've built.
00:17:38.319 --> 00:17:40.079
We can do this at industrial grade.
00:17:40.079 --> 00:17:41.839
Not that you can deliver the payments.
00:17:41.839 --> 00:17:44.160
I mean, that you know, you can just plug into someone and they'll do it.
00:17:44.160 --> 00:17:50.000
But the the whole sequence, like where does it start, where does it go, how does the statements work, how do you chase it down across six different providers?
00:17:50.000 --> 00:17:53.200
That's the muscle we've built in the last you know, two, three years.
00:17:53.680 --> 00:18:06.720
It's funny though, because you're also serving companies, many of which have been around for many, many, many decades and are probably pretty ingrained in their ways of doing banking, right?
00:18:06.720 --> 00:18:14.400
Hey, they're used to sending an ACH, they're used to sending a wire and waiting seven days and dealing with massive FX fees and everything like that.
00:18:14.400 --> 00:18:22.880
But I think that also gives you an immense opportunity to go in there and really, really disrupt the current status quo with a lot of these larger companies.
00:18:22.880 --> 00:18:34.880
I talked to a lot of people that are building quote unquote neobanks or payment rails for more modern tech-focused SMBs, but for the larger enterprises that have been doing this for 100 years, the inertia is strong.
00:18:35.039 --> 00:18:36.880
It's it's very, very strong.
00:18:36.880 --> 00:18:42.799
And and and also part of the value of in uh international, because really our sweet spot is international, it's not in the US, right?
00:18:42.799 --> 00:18:44.400
It's it is exactly that.
00:18:44.400 --> 00:18:51.440
People forget how hard it is to send a wire in a column like in the Colombian banking ecosystem.
00:18:51.440 --> 00:18:53.839
You have to actually go into the bank to send a wire.
00:18:53.839 --> 00:18:57.440
So the fact that we can we have an app where you can do it in the app, people will pay for that.
00:18:57.440 --> 00:19:02.160
Because like imagine if you're the CFO, it's like, do I go into a bank or do I do this in this app?
00:19:02.160 --> 00:19:05.279
And then everything is there, it's all in one screen, it's the same statement.
00:19:05.279 --> 00:19:09.759
Yes, there's a cost, but you know, there's differentiated value that's being provided.
00:19:09.759 --> 00:19:15.440
I would almost compare it some in some points to the uh the banking system in the US maybe 15, 20 years ago, right?
00:19:15.440 --> 00:19:16.640
Just in terms of the convenience.
00:19:16.640 --> 00:19:18.640
Doesn't mean like every country has that, by the way.
00:19:18.640 --> 00:19:21.200
Like Brazil's banking system is incredibly sophisticated.
00:19:21.200 --> 00:19:23.519
It's actually, I'd say, better than the US, right?
00:19:23.519 --> 00:19:25.039
It's faster, everything is instant.
00:19:25.039 -->
00:19:26.559
Like Brazilians are always shocked.
00:19:26.559 --> 00:19:27.759
Domestically, domestically.
00:19:27.759 --> 00:19:28.400
Domestic.
00:19:28.400 --> 00:19:34.960
When they do a payment and the last leg is the US leg, it's like, oh, it could come today, it could come tomorrow, it says 48 hours, and they're like, what does that mean?
00:19:34.960 --> 00:19:36.079
Like, why is it not instant?
00:19:36.079 --> 00:19:39.440
Like it's always confusing, but that's the value internationally.
00:19:39.440 --> 00:19:40.319
There is a gap.
00:19:40.319 --> 00:19:50.160
There's these strong banking players that like run the entire market, but like they there's no incentive to innovate because they it's it's a pure, amazingly profitable monopoly.
00:19:50.319 --> 00:19:52.960
So yeah, that's very, very fair.
00:19:52.960 --> 00:20:00.559
But at the same time, what you are doing is serving a need that will always be a need.
00:20:00.559 --> 00:20:04.559
As in, I don't think that any business is going to come to you and say, delete.
00:20:04.559 --> 00:20:06.799
I wish my payments payments were slower.
00:20:06.799 --> 00:20:09.359
I wish I had less countries that I could do payouts in.
00:20:09.359 --> 00:20:12.160
I I wish that my card programs had more friction.
00:20:12.160 --> 00:20:16.400
Like people always are going to want better, faster, cheaper when it comes to their money.
00:20:16.400 --> 00:20:20.640
Uh and if you give them a path for that, I think that's immensely valuable.
00:20:21.039 --> 00:20:23.519
Yeah, I so I fully obviously agree with that.
00:20:23.519 --> 00:20:28.880
I I think the harder part with enterprise is that it it is still a true enterprise sales cycle.
00:20:28.880 --> 00:20:30.799
So it's a 90-day, 120-day.
00:20:30.799 --> 00:20:33.359
You don't just show up and be like, hey, let's like test this out.
00:20:33.359 --> 00:20:34.720
It's like you gotta go through the process.
00:20:34.720 --> 00:20:41.440
You gotta and that's why it tends to be stickier, because once you do all that, then the bar is like, okay, is this better than what we had before?
00:20:41.440 --> 00:20:46.480
And like just like you're calling out, like that bar is not super high because you're competing with you know local banks.
00:20:46.480 --> 00:20:50.240
Um, but that cycle to get them to switch and try, that's actually really hard.
00:20:50.240 --> 00:20:53.519
That's not it's not like SMBs because SMBs are like, oh well, let me try it.
00:20:53.519 --> 00:20:54.480
Oh, 5% better.
00:20:54.480 --> 00:20:54.880
You know what?
00:20:54.880 --> 00:20:56.319
That's actually going directly to my bottom leg.
00:20:56.319 --> 00:20:57.519
Like, I'll test it out.
00:20:57.519 --> 00:20:58.400
That's not how it works.
00:20:58.400 --> 00:21:00.000
It works with like enterprises.
00:21:00.319 --> 00:21:00.559
Yeah.
00:21:00.559 --> 00:21:07.119
And what do you think is the major sticking points for the clients that are our core G's clients, right?
00:21:07.119 --> 00:21:11.839
You can do AR, you can do AP, you can do these cards, you can do all of these different things.
00:21:11.839 --> 00:21:19.039
Are there like one or two core features where companies, when they hear your pitch, they're like that, I I want that.
00:21:19.519 --> 00:21:21.759
So cards is still the opening wedge.
00:21:21.759 --> 00:21:27.519
And when I say cards, the big the biggest part there is the it's not just the cards that you spend, but the controls.
00:21:27.519 --> 00:21:30.960
Because at an enterprise level, the number one thing they care about is controls.
00:21:30.960 --> 00:21:32.880
Like, can I control my employee spend?
00:21:32.880 --> 00:21:37.599
If I give them a card, can it only work on an Uber and not work in this hotel or you know, for dinner?
00:21:37.599 --> 00:21:38.880
Like that's what they care about.
00:21:38.880 --> 00:21:40.000
That's what we do very well.
00:21:40.000 --> 00:21:41.839
We can do it well, we do it at scale.
00:21:41.839 --> 00:21:48.480
That's still, I would say, the biggest opener for us because it's something that is just so structurally different from a local bank.
00:21:48.480 --> 00:21:51.839
That's just not where they compete, that they don't care about it good or bad.
00:21:51.839 --> 00:21:53.359
Like that's not where they make their money, right?
00:21:53.359 --> 00:21:59.759
For us, that's that's the core opener, which is you come to us for controls, you figure out your card spend, then what we see is usually around.
00:21:59.759 --> 00:22:03.359
Month three or four, they start looking at, oh, cool, you guys also have payments.
00:22:03.359 --> 00:22:04.799
You know, maybe we'll try a little bit.
00:22:04.799 --> 00:22:10.319
And in the first 12 months, the revenue retention for our cohorts is almost 385%.
00:22:10.319 --> 00:22:17.839
So that means if you start with $100 of revenue, at the end of the month, end of the 12-month cycle, you're looking at almost $400 of revenue, which is very, very good.
00:22:17.839 --> 00:22:20.559
I mean, it just is a very sticky growing product.
00:22:20.559 --> 00:22:30.160
But cards, to your question, that's still the opener, and it usually sits around controls, which is how do I know that the card spin can only be done for ABC and it can't be done for XYZ?
00:22:30.960 --> 00:22:31.599
Interesting.
00:22:31.839 --> 00:22:33.839
I I want to talk about the future a little bit.
00:22:33.839 --> 00:22:37.839
I mean, obviously, Jeeves is on a legendary tear, right?
00:22:37.839 --> 00:22:44.079
You're raising hundreds of millions of dollars, you're seeing your revenue double, triple in a very, very short amount of time.
00:22:44.079 --> 00:22:47.680
What do you spend all of your day obsessing over?
00:22:48.799 --> 00:22:49.680
So many things.
00:22:49.680 --> 00:22:57.359
Um I think I think on a structural level, the good and bad of what we do is we cover a lot of surface area and we go pretty pretty deep.
00:22:57.359 --> 00:23:02.160
Like in Brazil, Mexico, Colombia, which is 65, 70% of our revenue, we're fully local.
00:23:02.160 --> 00:23:06.799
We have local entities, we're getting a local banking license in Brazil to take deposits, et cetera.
00:23:06.799 --> 00:23:11.440
So we are exposed to a lot of macro and regulatory tailwinds.
00:23:11.440 --> 00:23:12.880
Uh not tailwinds, just wins, right?
00:23:12.880 --> 00:23:15.359
It could be tailwinds, it could be whether it be tailwinds or headwinds.
00:23:15.440 --> 00:23:15.680
Yeah, yeah.
00:23:16.000 --> 00:23:17.200
Yeah, it's just it's wins, right?
00:23:17.200 --> 00:23:20.000
And something changes, it does have an impact on our business.
00:23:20.000 --> 00:23:21.759
A new, you know, election is coming.
00:23:21.759 --> 00:23:23.119
There's that that has an impact.
00:23:23.119 --> 00:23:26.240
Uh that there's, you know, new resolutions that has an impact.
00:23:26.240 --> 00:23:31.759
The flip side is out of all the businesses, we are actually probably most protected because we have 25 countries.
00:23:31.759 --> 00:23:37.279
So if one part of one country changes, we still, it's not, you know, the majority of the business.
00:23:37.279 --> 00:23:45.200
That's the flip side, but that's number one thing that keeps me up at night because there's just a lot of surface area like the war in the Middle East, right?
00:23:45.200 --> 00:23:47.119
We have companies that have fuel exposure.
00:23:47.119 --> 00:23:49.920
And our core product is still like a wrecks or a ramp.
00:23:49.920 --> 00:23:51.119
It's a 30-day charge card.
00:23:51.119 --> 00:23:55.119
That means for 30 days, we give you spend, and on day 31, you got to pay us back, right?
00:23:55.119 --> 00:23:57.920
And so there's a risk for those 30 days that we have to underwrite.
00:23:57.920 --> 00:24:01.200
And if you are exposed, you're an airline and you're exposed to fuel.
00:24:01.200 --> 00:24:03.359
I mean, that's a pretty big macro risk, right?
00:24:03.359 --> 00:24:05.519
Because your margins are now changed, right?
00:24:05.519 --> 00:24:09.680
So, like those are the it's almost like you have to be everywhere at all the time.
00:24:09.680 --> 00:24:15.359
It's not just small businesses, it's massive kind of enterprises with different levers that each one is working.
00:24:15.359 --> 00:24:25.440
So, what I see at Jeeves is you either come in and you usually know in one or two weeks it either becomes like super exhausting or it's almost like a exhilarating, like you just need more because it's like, you know, it just keeps you going.
00:24:25.440 --> 00:24:27.759
So that's the first one.
00:24:27.759 --> 00:24:32.880
And then the second one, I think this is the the way companies are building is also changing.
00:24:32.880 --> 00:24:40.079
And I'm always paranoid, I'm paranoid in general, but I'm always more paranoid that like there's going to be faster companies coming in than us.
00:24:40.079 --> 00:24:45.359
But the good thing with
fintech is exactly what you said, which is even though the tech is compressing, the fin is not.
00:24:45.359 --> 00:24:49.680
You still need regulation, you need you still need compliance, you cannot issue without a license.
00:24:49.680 --> 00:24:52.160
Those are things that are very defensible now.
00:24:52.160 --> 00:24:55.440
Almost the inverse of like four years ago where everything was the tech.
00:24:55.440 --> 00:24:57.039
It's like, oh, fintech, fintech, fintech.
00:24:57.039 --> 00:24:58.160
The tech is moving so fast.
00:24:58.160 --> 00:25:02.400
The cost of development, cost of software is so low that the tech is compressing.
00:25:02.400 --> 00:25:05.839
But the fin is one of the few industries that's regulated.
00:25:05.839 --> 00:25:06.720
Each country is different.
00:25:06.720 --> 00:25:09.039
You got to get a license, you've got to talk to the central bank.
00:25:09.039 --> 00:25:11.599
Like that, you can't AI your way out of.
00:25:12.240 --> 00:25:12.880
Yeah.
00:25:12.880 --> 00:25:22.000
It to your point there, I we talked to a lot of people who are also going and pursuing these licenses and talking to banks and everything like that.
00:25:22.000 --> 00:25:36.400
And I think that you know when you go and you look at the market and you see all of these new banks popping up all the time, there's oftentimes a really, really big discrepancy between what people are marketing and what people actually have on the back end and what they're actually building.
00:25:36.400 --> 00:25:40.799
You know, you see a bit of these websites, and it's like, oh, global money API, and oh, we're global.
00:25:40.799 --> 00:25:49.440
And the reality is, is oftentimes it's a wrapper of a rapper of a rapper of someone who is doing the work that you guys are are going out there and doing.
00:25:49.440 --> 00:26:06.160
And I think that the people who spend the time now are absolutely going to be paying dividends on that in the future, which sort of leads me to the next thing that I want to hear about is like when you guys are thinking about the future, what are you building towards right now?
00:26:06.160 --> 00:26:08.000
Like your engineering, your product.
00:26:08.000 --> 00:26:14.160
Is this, hey, we need to just keep moving up the stack and go deeper in AR and AP?
00:26:14.160 --> 00:26:16.400
Is it okay we need to make things agentic?
00:26:16.400 --> 00:26:19.519
What are you doing with with your $100 million round?
00:26:19.920 --> 00:26:22.160
So um so there's a few things.
00:26:22.160 --> 00:26:28.799
One, I think independent of uh whether it's agenc or not, we need to have ownership of the infrastructure.
00:26:28.799 --> 00:26:31.920
That's been our core philosophy since day one.
00:26:31.920 --> 00:26:37.759
We rent to start, we test the market, we see which ones are accelerating, and then we want to get the infra get the license.
00:26:37.759 --> 00:26:38.799
Like that doesn't change.
00:26:38.799 --> 00:26:45.599
So independent of like
what the interface is, because to me, agentic is the form of an interface, but the core rails, it doesn't change.
00:26:45.599 --> 00:26:50.480
It's cards, it's payments, you need stripe, you need someone that can have give you that access, whether it's a human or agent.
00:26:50.480 --> 00:26:52.000
So that is like number one.
00:26:52.000 --> 00:26:53.680
We have to go do these modules.
00:26:53.680 --> 00:26:54.640
That's not gonna change.
00:26:54.640 --> 00:26:56.799
We have to get the licenses we need to operate.
00:26:56.799 --> 00:27:02.559
Not every single one, you have to always like make that call, but the ones that we need to operate at the best cost that we can, like we have to get this.
00:27:02.559 --> 00:27:03.440
So that's step one.
00:27:03.440 --> 00:27:06.799
So there's a whole cycle and a whole team, like that's all they're doing, right?
00:27:06.799 --> 00:27:08.079
And so that's one block.
00:27:08.079 --> 00:27:18.079
The second one, I strongly, strongly believe uh that it's inevitable, probably in the next two to three years, that we are going to go fully into this era of like programmable money.
00:27:18.079 --> 00:27:19.599
And and there are two components of that.
00:27:19.599 --> 00:27:27.359
You have the agent side, which is an agent that has context enough that a CFO at an enterprise will trust that agent, which is still a pretty high bar, right?
00:27:27.359 --> 00:27:31.359
Because you have to understand CFOs are downside protection oriented.
00:27:31.359 --> 00:27:34.799
They don't care as much about the upside, but they care a lot about the downside.
00:27:34.799 --> 00:27:38.640
So you have to convince them that the downside is protected and then show them the upside.
00:27:38.640 --> 00:27:39.680
It's like, okay, cool.
00:27:39.680 --> 00:27:42.079
But if you can't convince them on that, you're not gonna get the first one.
00:27:42.079 --> 00:27:44.720
I think that box is still open, right?
00:27:44.720 --> 00:27:49.440
It starts with consumer, then you have the SMBs, then you have like it, it moves through the progression chain.
00:27:49.440 --> 00:27:54.000
But the enterprise is probably the last one in the B2B cycle because they move the most, right, to get to that point.
00:27:54.000 --> 00:27:55.279
So I think that's one.
00:27:55.279 --> 00:28:04.319
And then the second one is once you do have agents that are working without a human in the loop, the main thing is you have to be able to take money and break it up into very small pieces.
00:28:04.319 --> 00:28:07.839
And so this is where stablecoin to me comes into play.
00:28:07.839 --> 00:28:12.799
And I think there's components here that need to be figured out because today the cost can we can we pause here, Dalip?
00:28:13.039 --> 00:28:15.680
Why is money gonna be broken up into smaller pieces?
00:28:15.680 --> 00:28:18.240
Because that's a big statement.
00:28:18.559 --> 00:28:26.720
Yeah, so so to me, the way I see it is agents are going to be making money decisions at a frequency that is gonna be hard to comprehend for humans.
00:28:26.720 --> 00:28:34.480
You you're already seeing the outlines of this, where like I don't know which agent I have that's working now or not, but they're all they're all doing something, right?
00:28:34.480 --> 00:28:36.960
Like they're checking a reservation, they're checking a flight.
00:28:36.960 --> 00:28:47.920
And and that is only gonna keep increasing, which is they need access to spend, but they also need access to spend that is not necessarily in the same dollar multiples that humans operate in.
00:28:47.920 --> 00:28:53.119
Because for them, them I use very broadly for agents, the the value of time is very different.
00:28:53.119 --> 00:28:54.960
They're always on, they're always there.
00:28:54.960 --> 00:28:59.279
For human, like I'm not gonna spend five hours trying to find the best deal or like so.
00:28:59.279 --> 00:29:02.480
I'm like, it's almost like I have six how many transactions do you do in a day?
00:29:02.480 --> 00:29:03.519
Six, maybe.
00:29:03.519 --> 00:29:05.119
Like just in the full day, right?
00:29:05.119 --> 00:29:07.359
Like like you buy coffee, you go to restaurants.
00:29:07.359 --> 00:29:10.960
Yeah, like like it's you know, it's not that many.
00:29:10.960 --> 00:29:15.359
But I think an agent, at least from my thesis, is that number is gonna be 60.
00:29:15.359 --> 00:29:25.359
And if you get to 60, unless your budget as a human has changed, which is not gonna change, part of it is the cost and the frequency is gonna be tied to smaller dollar value transactions.
00:29:25.359 --> 00:29:32.400
And if you do smaller dollar value transactions, what we have not figured out, and we will, is how do you make the cost of that also lower?
00:29:32.400 --> 00:29:39.279
Meaning the swipe for a card still has a cost, an innate cost, whether it's one cent or one dollar.
00:29:39.279 --> 00:29:42.079
And at one dollar, it's priced in, like you don't really notice it.
00:29:42.079 --> 00:29:46.319
But if you get to one cent and lower, the cost of the swipe is actually more expensive than the one cent.
00:29:46.319 --> 00:29:51.039
So like there's there's a lot there that I like again, I don't think it's like insurmountable.
00:29:51.039 --> 00:29:53.200
I just think that is not something that's been figured out.
00:29:53.200 --> 00:30:00.400
But if you ask me in five years, I think even in three years, it's inevitable that that's the next frontier, which is this fully programmable money.
00:30:00.400 --> 00:30:08.160
You have the agents, you have humans in loop, but the agents are doing a lot of this and they need money they can break up, and then they need context enough that they don't f it up.
00:30:08.640 --> 00:30:20.640
So just to be clear, you believe that stable coins are going to be the currency of agents, and then agents are going to need to interact with stable coins because interchange and card fees are not going to cut
it.
00:30:21.359 --> 00:30:26.000
Because uh so I believe that yes, the first one, which is the stable coins are gonna be the currency for agents.
00:30:26.000 --> 00:30:28.960
And just to be clear, I don't think it has to be a USD denominated stablecoin.
00:30:28.960 --> 00:30:34.319
The concept of like digitized money you can break up into very minuscule components.
00:30:34.319 --> 00:30:36.400
That's what I think stablecoin provides.
00:30:36.400 --> 00:30:42.319
And then the second one, the reason is that I do think part of this agent economy is the frequency is going to change.
00:30:42.319 --> 00:30:48.160
I'll give you a almost like a non-agent version of this, but because we have this card product, which is a card API.
00:30:48.160 --> 00:30:52.400
But if you think about an API as an MCP, MCP is an agent tick, it's it's in the same box.
00:30:52.400 --> 00:31:01.440
The number of cards that companies were creating when we offered this thing, which is single use, we were talking about $100,000, $200,000, 200,000 cards a month.
00:31:01.440 --> 00:31:02.400
A month.
00:31:02.400 --> 00:31:04.240
Because it's fully single use, right?
00:31:04.240 --> 00:31:05.680
That's how an agent thinks.
00:31:05.680 --> 00:31:12.079
And and that's a fundamental difference from like how you and I operate, which you have one card or you know might have this card for movies, etc.
00:31:12.079 --> 00:31:13.599
So that's what's gonna start changing.
00:31:14.319 --> 00:31:19.519
I the mental model that I use is and there's a fierce debate, right?
00:31:19.519 --> 00:31:21.119
Are agents gonna use cards?
00:31:21.119 --> 00:31:21.599
Is that okay?
00:31:21.599 --> 00:31:23.200
Are agents gonna use stable coins?
00:31:23.200 --> 00:31:29.279
My mental model is agents are going to use cards to buy the things that humans would buy.
00:31:29.279 --> 00:31:37.200
If you want your agent to go and buy a plane ticket or a banana or a t-shirt, the agent can use a card, right?
00:31:37.200 --> 00:31:39.839
Agent fable cards, great, right?
00:31:39.839 --> 00:31:55.039
If you want your agent to go and use something that only an agent would use, compute inference, buy you know, some data, whatever it might be, stable coins are going to have to be the unit of currency specifically because of the reasons that you mentioned.
00:31:55.359 --> 00:31:58.160
Yeah, I I I I so one I fully agree.
00:31:58.160 --> 00:32:05.839
I think the third one that I would also throw out there that combines a little bit is we are still thinking about outcomes and tasks like humans do.
00:32:05.839 --> 00:32:13.200
And I think one of the things that will start changing is the sequence of tasks that you need to get the outcome might be different for an agent.
00:32:13.200 --> 00:32:16.400
So, like as an example, when we, and this is just human nature, right?
00:32:16.400 --> 00:32:18.400
We think about something, it's like five steps to get there.
00:32:18.400 --> 00:32:21.759
Then you're like,
okay, AI, I'm gonna use AI, and they'll help me with these two steps.
00:32:21.759 --> 00:32:23.440
And you're solving for the same outcome.
00:32:23.440 --> 00:32:28.720
But agents, and this is what to me is also gonna change, they might have a very different way of getting to the same outcome.
00:32:28.720 --> 00:32:40.160
And I I fundamentally believe you need access to payment reels, but you also need dollar components that you can break up because how they get to the outcome might be very different from how a human is thinking is the most efficient way to get to that outcome.
00:32:40.160 --> 00:32:48.559
And then they need the tools to be able to get to that outcome with the constraint of like solve this problem, not take these exact same six steps that I did.
00:32:49.680 --> 00:32:51.759
It's it's funny, we're talking very generically.
00:32:51.759 --> 00:32:53.119
I want to narrow it a little bit.
00:32:53.119 --> 00:32:56.160
Like you, these are your personal beliefs.
00:32:56.160 --> 00:33:00.240
How is Jeeves building for this agentic future?
00:33:00.240 --> 00:33:13.920
I mean, you know, when we talk about agents that the CFO of BMW and Lululemon and all these massive enterprises are gonna use, you guys are to, you know, to use your own words, the financial operating system, the financial operating layer.
00:33:13.920 --> 00:33:15.759
How do you build something for that?
00:33:16.799 --> 00:33:19.359
I think the reality there, and it's a good, it's a good push.
00:33:19.359 --> 00:33:23.519
The reality there is, so today we do have, you know, eight, eight, eight or ten agents.
00:33:23.519 --> 00:33:25.359
I'll tell you the two that people use.
00:33:25.359 --> 00:33:29.119
So we have one that actually um does GL coding.
00:33:29.119 --> 00:33:37.359
So if you're an enterprise, there's usually a person, it's a really tedious job that they look at every transaction and then they match it to like, okay, this is Uber, this is a restaurant, etc.
00:33:37.359 --> 00:33:38.640
It's called GL coding.
00:33:38.640 --> 00:33:41.920
Um immense value because you need that to close your books.
00:33:41.920 --> 00:33:44.960
We have a lot of data because we have companies that do this today for us.
00:33:44.960 --> 00:33:49.359
So we trained an agent, it's 99.99% accurate, so better than a human.
00:33:49.440 --> 00:33:52.000
So that's something is that data spending or reconciling?
00:33:52.400 --> 00:33:53.359
Reconciling.
00:33:53.359 --> 00:33:54.960
This is reconciling data.
00:33:54.960 --> 00:33:56.880
It's after the data comes back, they reconcile it.
00:33:56.880 --> 00:34:00.160
And then the this the this uh agentic card creation.
00:34:00.160 --> 00:34:01.759
That's another one that clearly is there.
00:34:01.759 --> 00:34:05.359
We have six other agents that that are not being used.
00:34:05.359 --> 00:34:15.440
And and to your point, I think part of it, and this is the reality of where we sit a little bit, is that one, internationally, the cost of human labor is actually lower, right?
00:34:15.440 --> 00:34:21.199
And so at some point you're fighting with like, okay, yes, maybe it's a little bit better, but I already have a team, they do it well.
00:34:21.199 --> 00:34:23.199
It's just one less thing I need to worry about.
00:34:23.199 --> 00:34:26.480
And like, you know what, let's just continue with this for now, and then we'll come back.
00:34:26.480 --> 00:34:31.519
And so the urgency of like you have to do this to agents, I I I think it's a very valid point.
00:34:31.519 --> 00:34:39.760
I I don't think it's there yet at a lot of the companies that we sell to because they have their process, everybody's busy, they have 10 things to do.
00:34:39.760 --> 00:34:44.159
And it's like, okay, give me the two things that work, and then I need to see real ROI.
00:34:44.159 --> 00:34:49.760
And then the rest is like, okay, good enough, but like not really pushing the needle of what I need, and so I'm not gonna invest in that.
00:34:49.760 --> 00:34:53.039
I I don't know if that's gonna change uh very quickly.
00:34:53.039 --> 00:35:00.559
Like we we don't, you know, like we operate internationally, and it's not San Francisco, like it's not like everyone's just going out and you know, saying, like, hey, if you don't do this, the world's gonna end.
00:35:00.559 --> 00:35:04.239
Like it's like those different origins of how that works, right?
00:35:04.239 --> 00:35:07.280
So I I think it's a it's a it's a very fair pushback.
00:35:07.280 --> 00:35:11.679
But but I I think that's still different from your first question, which is where you see this in three to five years.
00:35:11.679 --> 00:35:13.440
I do think we're gonna get to that point.
00:35:13.440 --> 00:35:16.000
But we have to build with that vision in mind.
00:35:16.000 --> 00:35:18.559
I think the usage, it's it's it's a very valid point.
00:35:18.559 --> 00:35:22.159
Like I think out of the six or eight agents we have, like two people actually use.
00:35:22.159 --> 00:35:24.880
And then six or eight, it's there, it's interesting.
00:35:24.880 --> 00:35:30.400
It's just not at the level of like, okay, I'm gonna go change my workload because of this one agent or two agents.
00:35:30.800 --> 00:35:31.119
Of course.
00:35:31.119 --> 00:35:39.840
And how much of what you do is actually in the flow of funds and moving money versus being the operating layer on top of that flow?
00:35:40.400 --> 00:35:42.880
Uh it's it's more the operating layer.
00:35:42.880 --> 00:35:48.480
We we don't touch the money directly unless we have the direct license, because once you do that, then you're in the funds flow.
00:35:48.480 --> 00:35:53.039
Uh so usually we'll have we'll have partners that can handle the fund delivery in and out.
00:35:53.039 --> 00:35:56.719
Uh I mean in Brazil, for instance, we are getting the license at the end of the year, so that'll change.
00:35:56.719 --> 00:36:04.480
Mexico, we are regulated so that
we can touch, but usually you don't want to be touching the funds at any point unless you have that direct license.
00:36:04.480 --> 00:36:07.440
So we see ours as like the orchestration layer.
00:36:07.440 --> 00:36:15.440
And also, people forget about this, but it's not just the orchestration, it's it's the value of the CFO not having to do that, which is they also could do this.
00:36:15.440 --> 00:36:20.960
They also could go talk to six providers, piece it all together, have like a Treasury ops team or FinOps team do it.
00:36:20.960 --> 00:36:22.639
That's what we sell, which is you don't have to do that.
00:36:22.639 --> 00:36:23.199
We'll do that for you.
00:36:27.679 --> 00:36:27.760
Yeah.
00:36:28.719 --> 00:36:29.039
100%.
00:36:29.039 --> 00:36:34.719
And what's really interesting, and this is a little bit of our secret sauce, is the companies we sell to look like Jeeps.
00:36:34.719 --> 00:36:36.159
We operate in 20 countries.
00:36:36.159 --> 00:36:37.360
We have 60 bank accounts.
00:36:37.360 --> 00:36:42.960
Every day I get a report, it's called cash in, cash out, which we're trying to automate, but it's still so hard to automate.
00:36:42.960 --> 00:36:49.039
Where someone literally goes into our 60 bank accounts, gets the actual cash that we have, puts it into a report, and sends it to me.
00:36:49.039 --> 00:36:52.320
So I know at the end of the day, our entire cash in, cash out, right?
00:36:52.320 --> 00:36:54.000
That's a manual, manual task.
00:36:54.000 --> 00:36:55.920
And now we're building agents to be able to do that.
00:36:55.920 --> 00:37:11.920
But what's good about where B sit is if our team uses it and they feel like it has value, I feel pretty confident we can sell that because the pain points we're solving is similar to what other organizations that operate in multiple countries with a sophisticated finance team will also have to solve.
00:37:12.639 --> 00:37:16.800
Yeah, it's like it's dog fooding your own product while they'll be able to do that.
00:37:16.800 --> 00:37:17.760
Which is unique, right?
00:37:17.760 --> 00:37:27.360
It's like you know, it's like I need to understand how to help CFOs that are building international finance companies.
00:37:27.360 --> 00:37:30.079
Which is and I'm gonna do that by building an international finance company.
00:37:30.320 --> 00:37:36.400
Yeah, by building it's and it's great because if our team uses it, then I feel pretty comfortable that we can package this and sell it.
00:37:36.880 --> 00:37:37.280
Absolutely.
00:37:37.280 --> 00:37:37.840
Absolutely.
00:37:37.840 --> 00:37:44.239
And I think you know, we we're sort of talking about the convergence of agents and and flow of funds and everything like that.
00:37:44.239 --> 00:37:55.440
But if you guys are able to develop agentic systems that are making a difference in your flows and in your back end and in your flow of funds, it's pretty easy to sell that, right?
00:37:55.440 --> 00:38:05.519
You sort of have that internal conviction on, hey, I know that this is going to bring you value because it's already done the and and this is how you use it, and here's how we use it.
00:38:05.599 --> 00:38:07.440
And you know, like like all of those help.
00:38:08.559 --> 00:38:20.000
That's that's really I I imagine as a CFO and you tell me if if I'm wrong, but like that level of trust is probably a really big boost to to your sales motion.
00:38:20.400 --> 00:38:31.199
Yeah, it's it's it's uh the the fact that we have the same learned experience is very helpful because we can sell it in a way that our team understands, meaning like it's useful for them.
00:38:31.199 --> 00:38:33.199
I I the trust is the most important thing.
00:38:33.199 --> 00:38:39.519
Like it's because to your point, like you we have this point of like the CFOs trust us, and now we're like, hey, try this with us.
00:38:39.519 --> 00:38:46.719
It's not that other companies don't have stablecoin, it's just that they don't have the entire thing integrated in a box that the CFO already trusts in these international countries.
00:38:46.719 --> 00:38:47.840
That's that's the reality, right?
00:38:47.840 --> 00:38:48.880
And why do they trust it?
00:38:48.880 --> 00:38:49.840
Because we have the infrastructure.
00:38:49.840 --> 00:38:50.559
Why do they trust us?
00:38:50.559 --> 00:38:51.599
Because they've used it for three years.
00:38:51.599 --> 00:38:55.519
Like it's all these little pieces, but it comes back to the same point, which is that trust.
00:38:55.920 --> 00:39:05.360
Yeah, I mean, I know a lot of companies that have really, really cool tech and think that that will do all of the heavy lifting.
00:39:05.360 --> 00:39:11.280
And I think most of the industry really underestimates the value of trust still.
00:39:11.679 --> 00:39:15.360
And and in Russia, and like the people around.
00:39:15.760 --> 00:39:16.079
Absolutely.
00:39:16.079 --> 00:39:27.599
I mean, but like historically speaking, you look back on the last like 5,000 years of finance, and the only through line and dictator of success is trust.
00:39:27.599 --> 00:39:28.639
That's it.
00:39:28.639 --> 00:39:36.719
And there's no reason why agents would throw that 5,000-year track record of trust away.
00:39:37.199 --> 00:39:38.159
It's it's money, right?
00:39:38.159 --> 00:39:43.599
It's it's like at the end of the day, it's like it's like you know, people get paranoid and irrational about money.
00:39:43.599 --> 00:39:44.639
I mean, that's just human nature.
00:39:44.639 --> 00:39:45.599
It doesn't change.
00:39:45.599 --> 00:39:46.480
I love it.
00:39:46.960 --> 00:39:47.360
I love it.
00:39:47.360 --> 00:39:47.840
I love it.
00:39:47.840 --> 00:39:54.800
So we started off the episode talking about your your big you know, hundred plus million dollar raise.
00:39:54.800 --> 00:39:59.920
I want to end the episode with talking about what's next post-raise.
00:39:59.920 --> 00:40:02.559
I mean, obviously, you guys are backed by the
best.
00:40:02.559 --> 00:40:05.599
I know your team well, you're building with the best.
00:40:05.599 --> 00:40:07.440
What can people expect out of Jeeves?
00:40:07.920 --> 00:40:08.639
So a few things.
00:40:08.639 --> 00:40:14.960
So one, um we are aggressively hiring uh and and aggressively for a very small team, right?
00:40:14.960 --> 00:40:16.559
So our team is still less than 200 people.
00:40:16.559 --> 00:40:19.199
We operate in 35 countries, we just added 10 new ones.
00:40:19.199 --> 00:40:24.079
But we are looking for more people, especially with depth on the stablecoin and crypto side.
00:40:24.079 --> 00:40:30.639
So what that means is we're looking for obviously folks on the product side, we're looking for engineering talent, we're looking for sales talent.
00:40:30.639 --> 00:40:40.239
We have a global head of partnerships role that is, I think, very exciting, which is you know, if you feel like you can do this across 35 countries, you'll have just what's unique about where we sit is we have a lot of levers.
00:40:40.239 --> 00:40:43.119
We have fiat, we have stablecoin, we have infrastructure, we have licenses.
00:40:43.119 --> 00:40:50.559
So you need someone that's creative to be like, oh, actually, we can mix and match this in a way that someone else hasn't done it, versus like here's a playbook that worked for XYZ.
00:40:50.559 --> 00:40:51.280
Let's take this here.
00:40:51.280 --> 00:40:59.679
So, one, what's going to change is we're moving into an era where we want to hire more because we want to 10x our surface area in the next year.
00:40:59.679 --> 00:41:02.320
And 10x means more geographies.
00:41:02.320 --> 00:41:03.920
It also means more product lines.
00:41:03.920 --> 00:41:07.280
So that's why we're also announcing the launch of the AR invoice suite.
00:41:07.280 --> 00:41:10.320
So we started with Brazil, but we want to expand that to all the countries we're in.
00:41:10.320 --> 00:41:15.840
Uh and then we also, I think we're mentioned this, but we also launched uh this global AI spend tracker.
00:41:15.840 --> 00:41:23.280
So our companies have this specific sequence where because they operate in six countries, they actually have AI spend in different currencies.
00:41:23.280 --> 00:41:30.800
So you have your engineering team spending in pesos with quad, and then you have your customer success team spending in pounds with OpenAI.
00:41:30.800 --> 00:41:34.880
And at the end of the month, you have the same question what is my spend and what is the ROI?
00:41:34.880 --> 00:41:37.840
And so we built this global uh spend tracker.
00:41:37.840 --> 00:41:40.400
It'll show it to you by model, it'll map it out for you.
00:41:40.400 --> 00:41:42.960
That's in beta right now, but we're gonna be launching that as well.
00:41:42.960 --> 00:41:44.880
So that's another area we're investing in.
00:41:44.880 --> 00:41:49.599
Uh and then we're also the the big announcement is like we are expanding to 10 new countries.
00:41:49.599 --> 00:41:58.159
So we've launched Argentina, uh, we're expanding into Peru, Paraguay, Uruguay, Central America, so that's Panama, uh, Dominican Republic, Costa Rica, et cetera.
00:41:58.159 --> 00:42:02.079
So this year is all Latam, and then next year we want to go into Middle East and then Southeast Asia.
00:42:02.079 --> 00:42:04.480
So we see our right to win kind of in this global south.
00:42:04.480 --> 00:42:05.679
So a lot to come.
00:42:06.480 --> 00:42:07.119
Fantastic.
00:42:07.119 --> 00:42:08.079
I love it.
00:42:08.079 --> 00:42:11.039
And and I'm I'm gonna be following along with that journey.
00:42:11.039 --> 00:42:13.280
I hope that our audience does as well.
00:42:13.280 --> 00:42:14.639
Dalip, thank you.
00:42:14.639 --> 00:42:17.039
Thank you so much for jumping on, for sharing your wisdom.
00:42:17.039 --> 00:42:21.760
You and I have about 10 more hours of questions to cover.
00:42:21.760 --> 00:42:25.679
Uh, but for the sake of the sanity of our audience, like I said, I appreciate you coming.
00:42:25.679 --> 00:42:28.400
For our audience, I hope that you were able to learn a lot.
00:42:28.400 --> 00:42:29.360
I know that I did.
00:42:29.360 --> 00:42:31.599
Uh, and Delete, we're gonna have to have you back on the show.
00:42:32.000 --> 00:42:32.480
I love it.
00:42:32.480 --> 00:42:34.559
I'm looking forward to continuing the conversation.
00:42:34.559 --> 00:42:35.920
Thanks for having me on.