Latest / The Index Podcast / Solana’s Agentic Finance Future with Bunny of Clawpump
Transcript
- 0:12SPEAKER_00: Welcome to the Index Podcast.
- 0:14SPEAKER_00: I'm your host, Nikki Scans.
- 0:16SPEAKER_00: Today I'm joined by Mauricio Trujillo Ramirez, co-founder of Claw Pump, the House of Agenc Finance and Tectonic, the Railway Builders for Institutional Agentic Commerce, known online as Bunny.
- 0:27SPEAKER_00: He's building the financial infrastructure for the age of autonomous AI.
- 0:31SPEAKER_00: Mauricio, welcome to the index.
- 0:33SPEAKER_00: And I guess before you even respond, you want to go by Bunny for the rest of this?
- 0:37SPEAKER_00: Or Mauricio, you've got a nice online presence under the bunny name.
- 0:41SPEAKER_01: I think Bunny is great.
- 0:43SPEAKER_01: I like being able to just tell people, hey, this is Mauricio also, because we'll we'll dive into this a little bit more in on the interview, but having having to work with both institutions also just like general like mass uh adopted like users.
- 0:58SPEAKER_01: I I like to also go by bunny and both my legal name.
- 1:03SPEAKER_01: So bunny is good for the advantage.
- 1:04SPEAKER_00: Bunny is better for mass adoption or Mauricio?
- 1:07SPEAKER_01: I think Bunny is better for mass adoption, just from the standpoint that it's something needs to pronounce.
- 1:11SPEAKER_01: Like let's say like no no one knows the founder of Jupyter's name, they just know him as Meow, right?
- 1:17SPEAKER_01: Yeah, and I think that's great.
- 1:18SPEAKER_01: I think uh being able to lean into that is amazing.
- 1:21SPEAKER_00: Where did the name originally come from?
- 1:24SPEAKER_01: Uh I'm a huge fan of Bud Bunny, and back in like 2019, 2020, I was mostly in financial Twitter actually, and then I realized that there were a lot of really large asset managers um that just had names like I I won't dog someone, but let's say one of them that manages like quite a few billion dollars, his ad on Twitter is Gabaghool, right?
- 1:45SPEAKER_01: Because he he just lost like that's in both cases his DJ name and also just his Italian-esque Gabaghool name, but yeah.
- 1:55SPEAKER_00: I love it.
- 1:56SPEAKER_00: So nothing too specific.
- 1:57SPEAKER_00: I I mean it's a cool, it's a cool background, and it for sure is memorable, and it's probably why you've been able to build up quite the online presence.
- 2:04SPEAKER_00: Do you feel like under you know, under the name Bunny, like why do you think you have been able to grow a little bit of a following, even outside even outside being a founder?
- 2:15SPEAKER_01: Um, I think you know, I I like to make people laugh.
- 2:19SPEAKER_01: I like to joke around on Twitter, I like to be able to then also I think that the ideal distribution strategy is kind of like make a joke on Twitter that just goes mega viral and increases the surface area of who your audience is, and then post like a very serious, like, oh, here's my company announcement post, and then everyone that already likes your jokes is gonna also see that and be like, oh my god, this person cannot only joke but also you know actually build something crazy.
- 2:44SPEAKER_01: So I think I that's why I've been building an audience.
- 2:47SPEAKER_01: I also like to play around uh outside of my ecosystem, right?
- 2:50SPEAKER_01: Like I I went from Fin tweets to like financial Twitter all the way to like I guess like tech Twitter and then eventually crypto Twitter, and now I'm on AI Twitter as well.
- 3:00SPEAKER_01: Um and now I guess I also am starting to enter like generative or like techno art Twitter, I guess.
- 3:08SPEAKER_01: You know, there you just gotta play around with people, and I think that that's how you build an audience here.
- 3:13SPEAKER_00: Yes, I I think that's like honestly, for any other founders out there, I think that's like really informative.
- 3:18SPEAKER_00: I mean it it leans into like the whole founder-led marketing narrative, but that doesn't mean that everything you post has to just be about your company or be super serious.
- 3:26SPEAKER_00: Like just be yourself, have fun.
- 3:27SPEAKER_00: Your audience will grow to love you, which means that they'll love what you're also building.
- 3:31SPEAKER_00: So I think I think you're doing a great job leading in that space and being an example.
- 3:35SPEAKER_00: So, bunny, something fun about this conversation is usually we only get to talk about one company, but you're already we you're already the founder of two companies, and we're gonna talk about both of them today, being claw pump and tectonic.
- 3:50SPEAKER_00: Please just break down for the audience the difference between the two, the similarities, and why, you know, why you are starting two companies at once.
- 3:59SPEAKER_01: So claw pump, the simplest thing I'll put it in is the house of agentic finance, right?
- 4:04SPEAKER_01: Like we very much uh are following on hyperliquist step on becoming the they claim to become the house of agency finance.
- 4:11SPEAKER_01: Of course, we believe that they are making strades on that.
- 4:14SPEAKER_01: Like they're pushing the pushing the envelope of what can be done, what is possible to be done on chain.
- 4:19SPEAKER_01: I think Solana is very much following after that as well.
- 4:22SPEAKER_01: And if anything, we want to help turn Solana into the house of agency finance, right?
- 4:26SPEAKER_01: Like we're building tools for agents to be able to trade on-chain, do prediction market arbitrage, do cross-perpetual arbitrage on-chain, right?
- 4:35SPEAKER_01: Like all of these super complex strategies that would traditionally be available to not even just only Wall Street, but super hyper-specialized uh crypto trading firms.
- 4:45SPEAKER_01: We're making them public and democratizing them for everyone, right?
- 4:48SPEAKER_01: And we don't want people to understand these strategies necessarily.
- 4:51SPEAKER_01: We want to help them use their agents to understand these strategies and just automate their trading in, automate their finances in the way that is simplest for them, right?
- 5:00SPEAKER_01: Uh so that's on the cloud on side of things.
- 5:02SPEAKER_01: And then something that we realized very early on the uh when we started working on cloppam was that there were a lot of the tools that we were making that actually also had institutional use cases.
- 5:16SPEAKER_01: So that is how eventually Tectonic is effectively like the birthday or like this byproduct of Cloud Pump that has a whole other institutional audience.
- 5:27SPEAKER_01: Like with Tectonic, we've collaborated with the Sonna Foundation, we've collaborated with Google, we've collaborated with now there's like a perpetuous exchange that wants us to build their agentic financial layers so other agents are able to plug into them and also allow, oh, and also even this other fund, like a traditional, just like general like Wall Street fund that wanted some help optimizing their agentic trading tooling as well.
- 5:54SPEAKER_01: Um so yeah, like now under the umbrella of Tectonic, we do all of these much larger agentic uh payments for financial integrations.
- 6:03SPEAKER_00: Okay, so you would say that claw pump is more like user-facing, retail facing, and tectonic is more like your institutional branch.
- 6:14SPEAKER_00: Correct.
- 6:14SPEAKER_01: Yeah, like right now the tectonics institutional branch, that is who we um they you can go to a Fortune 500 company and say, hey, I'm claw pump, right?
- 6:23SPEAKER_01: But you can definitely go and tell them, hey, I'm tectonic, and we've already secured like or like we've optimized trading infrastructure for two, five, ten billion dollars of asset center management, right?
- 6:36SPEAKER_01: So that's I I'm I like being able to um shift between both ecosystems, um, but of course, I think the having this separation of of church and state of the consumer facing and then the institutional facing has is playing nice for us right now.
- 6:52SPEAKER_00: Okay, what would you say like outside of the name sounding you know more professional and less retail facing, like what are the different demands that like institutions would have with tectonic versus you know, retail users would have with claw pump?
- 7:07SPEAKER_00: Like don't they have the same agentic needs or no?
- 7:10SPEAKER_01: They don't have the same ones, uh especially because uh okay, so this is where I think it's interesting, right?
- 7:17SPEAKER_01: So an average user does not necessarily care.
- 7:20SPEAKER_01: They they care about let's say like latency just from the sample of like I want my the user experience of pressing trade or pressing swap or telling my agent, hey, please swap this, and it being fairly fast, right?
- 7:31SPEAKER_01: So we do do a lot of very crazy back-end optimizations on being able to geographically locate your PC next to like whatever like battery or like whatever like data set you're calling.
- 7:43SPEAKER_01: Uh, we do all of these optimizations for the just for the nice UX for um average like retail user.
- 7:51SPEAKER_01: Uh, but then for an institution, we take that one step further and we tell them, let's say, let's say a large like 14500 company doesn't want to plug into a facilitator, right?
- 8:03SPEAKER_01: They want to own their own facilitator for X402 payments, and they are not gonna process just like a couple of transactions, right?
- 8:10SPEAKER_01: They're not gonna um process like a couple thousands, they're gonna process millions per second, and they're not gonna process millions per second on just one region, but across many different regions.
- 8:20SPEAKER_01: So that is where we will say, oh, okay, we built this infrastructure and now we're gonna make it horizontally scalable by deploying, having one-click deploy of this infrastructure across many different regions, right?
- 8:31SPEAKER_01: Um and that those types of lessons we can take back to the retail user and just empower them with these solutions, but only once they need them, right?
- 8:39SPEAKER_01: Like no retail user cares about like uh having geolocated access to some data that only institutions may need.
- 8:49SPEAKER_01: Uh, we want to make sure that we're we're over-engineering solutions that do lead to good user experience on CloudPomb, but then on the institutional side, we you know we want to be SL SLC2 compliant.
- 9:03SPEAKER_01: Uh we want to make sure that we give them proper telemetry so they can check the transaction at every single level of the stack and all of that.
- 9:10SPEAKER_00: Okay, I think that was a really good description.
- 9:12SPEAKER_00: Uh now as we go through things, I this is obviously someone viewing this, they might be learning a lot about agentic payments as we go here.
- 9:20SPEAKER_00: Um you brought up two terms.
- 9:22SPEAKER_00: One is like X402.
- 9:24SPEAKER_00: Um I'd love to hear you know how you view X402 as in like a currently the state of it, what the future looks like for it, and how it's like applicable to retail and institutions, or if it's only applicable to institutions.
- 9:36SPEAKER_01: Yeah.
- 9:37SPEAKER_01: So we have two, we have a couple views on that.
- 9:42SPEAKER_01: Um let's say just to take a concrete example.
- 9:45SPEAKER_01: On Claw Pump, right now, if you spin up an agent, uh, the agent is able to earn money through DeFi and pay for its own compute through an XRO2 gateway.
- 9:56SPEAKER_01: So with that, we've invented this term called the eternal agent.
- 9:59SPEAKER_01: So it's an agent that can just live on forever because the more as long as it can continue making money and it can continue on auto-improving the skills, their DeFi skills that they have access to, they should be able to live forever, right?
- 10:12SPEAKER_01: As long as they can continue compressing their memory and continue just feeding themselves back that memory once their their context video has gotten too bloated, all of that, they they should be good.
- 10:22SPEAKER_01: They you can use that agent effectively for as long as whatever these are as long as these services uh stay alive, right?
- 10:29SPEAKER_01: And even you can switch LMs and whatnot.
- 10:31SPEAKER_01: So I think that there's very pragmatic use cases of X42 that no other, let's say, an institution wouldn't do.
- 10:37SPEAKER_01: Like OpenAI could never release an internal agent, right?
- 10:40SPEAKER_01: Like that would be an insane headline, and they will get into a lot of weird uh I think they will get a lot of pushback, right?
- 10:47SPEAKER_01: Like the AI safety people will go crazy.
- 10:50SPEAKER_01: And even a traditional think that cannot even do these experiments with X402 because they also don't understand this infrastructure just yet.
- 10:57SPEAKER_01: Um so that's on the consumer side, I think that there's some good programming use cases.
- 11:01SPEAKER_01: Um on the institutional side, I think that X402 is, and it's gonna sound maybe a little bit weird for me to say, but it's overhyped.
- 11:12SPEAKER_01: I think that there's a lot of teams, uh there's a lot of institutions that signed up to the X402 Foundation and with the Linux Foundation and all of that.
- 11:21SPEAKER_01: And they are all talking about how the future of agenc payments uh will, sorry, the future of finance will be agentic payments, but they're not doing the agenc payments.
- 11:29SPEAKER_01: They're actually not going on the gateways, they're not uh they're not using them.
- 11:33SPEAKER_01: They are maybe building like a proof of concept, and that's about it, but they're not actually running them.
- 11:37SPEAKER_01: There's very few, like uh, like I count them on my hand, like institutions that are actually building very novel stuff with X402 and doing maybe they are still making some demos or prototypes or proof of concepts, just so then they can use that as a as a way to throw your course X402 into their actual stack.
- 11:57SPEAKER_01: So I think that there is 100% a future there uh where one day you should be able to pay for your let's say your database uh spent purely through X402, no Stripe sign in, no like link, no uh not even like oh sense tablecoin to here, just an agent takes care of that billing system for you, and maybe there's like a larger billing agent within an organization that just handles all of these payments, right?
- 12:25SPEAKER_01: Um I think that is where the X42 will have a future, but first we need to be sincere and actually take those initial steps instead of just saying that the the future will be for agentix payments, right?
- 12:38SPEAKER_01: Like no, there's very few teams actually shipping stuff.
- 12:41SPEAKER_01: And it's it's baffling to me that there's teams that have raised so much money and multi-billion dollar companies that have signed up to the to the foundation and only you know, like a startup of two people with the with the support of Google was able to already ship more than a lot of the foundation, right?
- 12:59SPEAKER_01: Like that's and that's not to tune our own corns, right?
- 13:01SPEAKER_01: But it's more so to say, like, hey guys, like you you should all be putting the legal work, because it's not engineering.
- 13:07SPEAKER_01: Engineering part is super easy, actually.
- 13:09SPEAKER_01: Uh the legal work and the PR work and the corporate uh politics work is where actually a lot of the the holdups exist.
- 13:20SPEAKER_00: Yeah, well that makes sense.
- 13:21SPEAKER_00: And for like a leaner team, you can sort of operate quicker within you know within those guardrails.
- 13:27SPEAKER_00: 100%.
- 13:28SPEAKER_00: Yeah.
- 13:28SPEAKER_00: So, Bunny, what I want to try to do is that we're gonna try to get a nice clip here.
- 13:33SPEAKER_00: Um, and I think a good way to get one is the X402, regular people like myself are seeing it all over the place as a term, and obviously you you were just describing how it, you know, how it relates to tectonic and and claw pump.
- 13:46SPEAKER_00: But for everyone else out there who's seeing it all over the timeline, not fully understanding it, we're gonna ask this question, hopefully get a nice clip from it.
- 13:55SPEAKER_00: How would you describe X402 and what it does and why it's Solana relevant?
- 14:03SPEAKER_01: The way I will describe X402 is uh agentic on-chain and off-chain handshake to give an agent off-chain data.
- 14:11SPEAKER_01: That's the simplest way of putting it.
- 14:12SPEAKER_01: Like maybe you can also describe it as a receipt of just saying, oh, did this uh did this entity make a payment to us?
- 14:20SPEAKER_01: Yes.
- 14:20SPEAKER_01: Okay, here's the data back, right?
- 14:22SPEAKER_01: The reason why it's relevant to Solana is because it increases the possibilities of what types of programs can be built on Solana, right?
- 14:31SPEAKER_01: Like before X402 existed, we couldn't have created eternal agents, right?
- 14:36SPEAKER_01: The eternal agents will have been um a fake marketing term in which an agent will have paid our wallet and then we will have used a credit card to pay their data bill.
- 14:47SPEAKER_01: Whereas now we're actually creating an X402 gateway for sorry, we have created an X402 gateway to let agents pay for their own compute, right, directly without having to go through us in a permission manner necessarily.
- 14:59SPEAKER_01: Um I think that X4 I think that is where the most fascinating things on X402 exist.
- 15:04SPEAKER_01: Um what can you let an agent pay for in the real world that they couldn't have done so before?
- 15:12SPEAKER_00: We had to take a few steps backwards um and just get more like uh you know, 50,000 feet view here.
- 15:18SPEAKER_00: Now, obviously, you know, with Claw Pump and Tectonic, you're bullish agentic payments, right?
- 15:24SPEAKER_00: Um, do you see AI agents sort of replacing the wallet interface on crypto?
- 15:30SPEAKER_00: I am very bullish agenc finance.
- 15:34SPEAKER_01: I am not really bullish on agenc payments for the most part, uh, as an economic process.
- 15:40SPEAKER_01: I am bullish on it as in like it'll be part of the day-to-day of operations, but I don't think that there's actually money to be made there.
- 15:48SPEAKER_01: I think that most facilitators will the margins will like crunch, they'll go to zero.
- 15:53SPEAKER_01: Uh, I think that a lot of the margin will be on the compliance side of things, um, on the different telemetry and whatnot, like DevOps things that you will have to do as a facilitator, right?
- 16:02SPEAKER_01: Like we've had to build so much custom infrastructure for Google uh for their facilitator that was insane.
- 16:11SPEAKER_01: Um I don't I've never built a product for like a proper institution until actually no, that doesn't lie.
- 16:16SPEAKER_01: I have worked on products for institutions, but as soon as you enter the payments realm for an institution, whole other legal department gets involved, whole other pre PMs, like uh the product manager is just like on you like every single day.
- 16:31SPEAKER_01: And we've been working with a team in Singapore, so I've been doing double shifts.
- 16:34SPEAKER_01: Like we work in the morning support club and then late nights from New York to uh on Singapore time zone.
- 16:41SPEAKER_01: Um but yeah, just to go back to the main thing, uh I the way I see it is that agents will take over a lot of the very traditional payments, uh, but I don't think that the current implementation of X42 makes sense.
- 16:53SPEAKER_01: I think that a lot of the tooling that agents will end up using later on is maybe, oh, I don't want to get I don't want to wait for my user to get this API key or get KYC on this thing.
- 17:04SPEAKER_01: I'm just gonna buy a little bit of this data or like a little bit of this service that I need for my own process to continue, and that's it.
- 17:10SPEAKER_01: Right.
- 17:11SPEAKER_01: I I think that is where I'm very bullish on.
- 17:15SPEAKER_00: Okay.
- 17:16SPEAKER_00: So do you view that you know crypto will be in like a post-human post-wallet world, or do you think no, it's gonna be a combination of the two?
- 17:25SPEAKER_01: Oh, I think sorry, yes, that um I think not even in the future, but like already, you're seeing that there's a lot of people that interact with crypto through the terminal, and there's other people that interact with crypto through their wallet, right?
- 17:38SPEAKER_01: Or not even through their wallet, but just through an app that just so happens to have a wallet as well.
- 17:43SPEAKER_01: So I think that that's how effectively it will continue to get split up.
- 17:48SPEAKER_01: Uh it's like this this bimodal distribution of are you a developer that needs crypto or are you a user that just so happens to also need uh crypto to be able to continue on whatever service to procure whatever services they want.
- 18:03SPEAKER_00: Okay.
- 18:04SPEAKER_00: And is there basically how our interactions going to, you know, how are interactions going to change as like the agents become a bigger part of all this versus it all just being humans on wallets, right?
- 18:17SPEAKER_00: So interfaces are gonna have to change alongside that, you know?
- 18:21SPEAKER_00: I I think so.
- 18:23SPEAKER_01: This is where I I wish I had a more shocking answer, but I think you literally won't change that much.
- 18:30SPEAKER_01: I think that every single developer out there that has already learned about X402, they have already changed their their stack in a way.
- 18:39SPEAKER_01: Like I said, like I I made this like joke uh prank phone call to a friend and use an X402 gateway to pay for the to pay for someone to spin up a temporary like phone and then to spin up a temporary like voice service uh to just like prank call my friends, right?
- 19:00SPEAKER_01: And I did that all through the terminal.
- 19:02SPEAKER_01: Uh and you you can do that through Cloud Pump as well, which I think is really funny.
- 19:06SPEAKER_01: But uh I think that me as a developer and other developers in the ecosystem, we've we've already changed the way we operate, right?
- 19:14SPEAKER_01: Like the the thing that has changed there, the thing that has quintessentially changed there is that I no longer have to learn all of these services and go, oh, I need to get an API key for this, an API key for that, I need to feed this context documents to this.
- 19:28SPEAKER_01: I can instead just like tap into any extra to marketplace or bazaar or whatever directory you want to call it.
- 19:37SPEAKER_01: Um, and then ask, hey, I want to do this process.
- 19:41SPEAKER_01: What are the best services that I can interact with?
- 19:42SPEAKER_01: And I can pay for them right now, right?
- 19:45SPEAKER_01: That is the UX.
- 19:46SPEAKER_01: It works right now.
- 19:47SPEAKER_01: I think that most of the services are not pushing the envelope on what actually can be is possible with X402.
- 19:54SPEAKER_01: Uh, but you know, that's that's how things tech always starts, right?
- 19:58SPEAKER_01: Like we first put something safe, then there's a bunch of experimentation, then we reel things back in and say, okay, these are the actual useful things that uh can be done.
- 20:07SPEAKER_01: Um and then for the consumer, I think literally nothing will change.
- 20:11SPEAKER_01: I think that the consumers, uh, I think the app developers still know that what is good for the consumer is just a nice user experience, right?
- 20:20SPEAKER_01: Open claw had a huge viral moment, and every single developer and maybe some people that are app builders wanted to include open cloud within their services, but that that hasn't led us anywhere just yet, right?
- 20:34SPEAKER_01: Like no uh Open Claw virality is like at all time lows right now, or it's near all-time lows.
- 20:40SPEAKER_01: Uh, people are not really talking about it anymore.
- 20:43SPEAKER_01: Um yeah, I think that uh for the consumer, we will always want a nice app.
- 20:49SPEAKER_01: Uh, but for the developer or like tech heavy or tech savvy power user, yes, they'll 100% win X402 within their terminal interface.
- 20:59SPEAKER_00: Bunny, I actually tend to agree with you.
- 21:01SPEAKER_00: And it's you know, it's funny talking to someone who is living in like this agentic world, you know, with your companies, but I I agree that a lot of what we've been told has been overplayed quite a bit, that like our entire experience is going to be you know changing and becoming agentic, and we're not going to have to use any apps on our own, or like it's just that's sort of betting against human experience, which I wouldn't bet against.
- 21:25SPEAKER_00: People love ordering things themselves and going onto a smooth, you know, UI UX.
- 21:32SPEAKER_00: Yeah.
- 21:34SPEAKER_00: No one is that no one is replacing no one is replacing that with an agent.
- 21:37SPEAKER_00: Like that is something set for a thousand years that people have loved and will continue to love.
- 21:42SPEAKER_00: And I would rather use a smooth UI and UX.
- 21:45SPEAKER_00: It may be agents can help improve, but to say that we're gonna replace all that altogether is unlikely.
- 21:52SPEAKER_01: Yeah, like I would hate to go and um like I I I admire Nikita Vir a lot, and let's say like a lot of He has this one super famous tweet about how every single app is about either making money or getting laid.
- 22:06SPEAKER_01: And I would hate to go on a Tinder day that was arranged by my agent.
- 22:11SPEAKER_01: Like I if I don't know what the person looks like, what they're into, like if I don't see an SUI of like, oh, like this is what the person is like, I would hate to have a person reduced to a JSON.
- 22:22SPEAKER_01: Right?
- 22:23SPEAKER_01: For the money side of things, maybe I don't mind if I go my bank, my bank's app and I say, hey, uh, make a little bit more money, right?
- 22:32SPEAKER_01: Like, or be a bit more aggressive uh while trading or in this market.
- 22:36SPEAKER_01: Maybe that's okay.
- 22:37SPEAKER_01: Um, but then to order a smoothie, as you said, uh to order my coffee.
- 22:42SPEAKER_01: Like I like walking in.
- 22:43SPEAKER_01: I went to a coffee shop earlier today.
- 22:44SPEAKER_01: I will hate it if I order uh my coffee through my a terminal interface.
- 22:49SPEAKER_01: Like I that just feels so dehumanizing to me.
- 22:52SPEAKER_00: Yeah, it's unlikely, bunny.
- 22:54SPEAKER_00: And we we've we've already seen so much like new technology come that and people are still doing things in older, outdated ways.
- 23:00SPEAKER_00: I mean, a perfect example is like, you know, a lot of a lot of like wealth management and like paying, you know, paying fees to wealth management when you have wealth management at your fingertips.
- 23:11SPEAKER_00: And what's been great about you know the unlock for AI is knowledge is now free, right?
- 23:17SPEAKER_00: So you have all the same data that anyone else who could be advising you would have, but still people choose to have that human connection around their money.
- 23:26SPEAKER_00: Like, and I think we're gonna run into a similar thing here where it's like, oh, we could move everything to agents, but we don't want to and likely won't.
- 23:34SPEAKER_01: Yeah.
- 23:34SPEAKER_01: And on top of all that, like agents, there's the old-timey story of oh, this they just have no taste and whatnot, right?
- 23:44SPEAKER_01: Like, and you we do keep encountering that every single time.
- 23:48SPEAKER_01: Uh and I say we as like the larger developer community that is very AI pilled or is on deep AI psychosis or and whatnot, right?
- 23:57SPEAKER_01: Um, you still need a lot of human guidance for a lot of things.
- 24:00SPEAKER_01: And even then, like people just like the the curation of things, right?
- 24:04SPEAKER_01: Like the non-GMO market is a thing, like the organically made, the handcrafted.
- 24:09SPEAKER_01: Um the there's like even this whole term called friction maxing that went viral on TikTok a bit ago, uh, about how people are remove adding back friction to some daily parts of their life because they feel like things have been made too convenient for themselves.
- 24:25SPEAKER_01: Um, I don't personally believe in that, but I also know that that entire side of the market exists, right?
- 24:30SPEAKER_01: Like people will still want to chat with a human just to get reassured because for some reason they don't trust technology for procedural things, even, right?
- 24:40SPEAKER_01: Like they they whenever something can have a taste variable or a cultural variable or uh a variable where you trust the conveyor or the conjugate of information more, that is where people will always choose humans, I think.
- 24:56SPEAKER_00: Yeah, I I definitely agree.
- 24:58SPEAKER_00: And you know, you're bringing up the the whole taste argument with AI is obviously super prevalent right now.
- 25:05SPEAKER_00: Um you're seeing a lot of you're finding that the individuals who, especially on the design side, the individuals who create the best front-end design uh using AI are designers by trade.
- 25:19SPEAKER_00: So that's sort of like the irony of it.
- 25:22SPEAKER_00: And I know a lot of um I know a lot of you know the the uh you know technical founders are are loving like that they can now do design with AI and it's great, it's a huge unlock for them.
- 25:34SPEAKER_00: But the reality is that like a designer using these tools will outperform them a hundred out of a hundred times with the same tool set.
- 25:42SPEAKER_01: Every single designer I know has built way better, way sicker stuff than I have ever built with AI, and I'm okay with that, and instead I just choose to sit down, learn.
- 25:54SPEAKER_01: I I feel like the throughout my my entire career, I've always been very self-critical and very cognizant of what are my limits.
- 26:03SPEAKER_01: I've definitely like before it was a lot of like BD stuff and like commercials, and I hit that ceiling and then I broke through it, then it was engineering.
- 26:12SPEAKER_01: Now I like especially like backend engineering.
- 26:15SPEAKER_01: Now we've definitely gone past that ceiling.
- 26:17SPEAKER_01: Like we've I built stuff like backend infrastructure for multi-billion dollar institutions that I I always wanted to work with, but you know, I never really had the skills to do so, and now now I can, right?
- 26:31SPEAKER_01: Like now I've been doing this for quite a few years.
- 26:33SPEAKER_01: Uh and now I feel like design is the next limit.
- 26:36SPEAKER_01: And I really I hope that in two years from now I'm able to see my work now and say, okay, this was slow, and now I can actually make something very pretty, right?
- 26:45SPEAKER_01: Thankfully, my my co-founder is a really good full stack engineer, but also a great designer, so I'm I'm very happy with that.
- 26:51SPEAKER_00: We all know it when we see it.
- 26:53SPEAKER_00: You can the speed, the efficiency, and then the result when someone with a design brain creates something.
- 27:00SPEAKER_00: There's something magical about it that when you see it, you're like, oh, like this person's talented at this.
- 27:07SPEAKER_00: Um and it's cool that you can recognize it, but also be like, I think I can achieve it eventually too, with like enough, you know, with enough practice.
- 27:15SPEAKER_00: Um I've gotten us way off topic, but this is all conversations that I really that I really love to have.
- 27:20SPEAKER_00: So I appreciate you humoring me on this.
- 27:22SPEAKER_00: Um to try to get us back on talking more claw pomp tectonic.
- 27:27SPEAKER_00: Um we it was brought up the you know, the cold start problem for agents.
- 27:31SPEAKER_00: Uh can you sort of explain in simple terms what that is and why it's a blocker?
- 27:36SPEAKER_00: For sure.
- 27:37SPEAKER_01: So the what I I call it a cold start problem because back in it hasn't been that many months, actually.
- 27:44SPEAKER_01: I I was gonna say like back in like 2020 something, but no, it literally like four months ago.
- 27:49SPEAKER_01: Four months ago, uh I okay, wait, let's back up a little bit more.
- 27:55SPEAKER_01: About two years ago, when I was just starting to use some agentic tooling to build like some on-chain stuff, uh I always did like the off-chain work with agents, and then as soon as it got to on-chain stuff, that is where I had to go in and do like some very nice organic, handmade code, right?
- 28:14SPEAKER_01: Uh so for that I was like, okay, well, maybe agents will get more context on on-chain building someday, right?
- 28:22SPEAKER_01: Then things progress really nicely.
- 28:24SPEAKER_01: Now, most agents do have uh some awareness of how to build on certain on-chain systems, uh, primarily VMs, I will say, because of purely because of the relationships that so many VM teams have with OpenAI and whatnot.
- 28:39SPEAKER_01: But then uh about four months ago, when people were looking at how to let an agent build on Solana, one of the complaints that we heard from people across the ecosystem and even within the Solana Foundation itself was that yes, agents are cool for on-chain development or on-chain operations.
- 28:59SPEAKER_01: However, they are always born without gas, right?
- 29:02SPEAKER_01: Uh so that's what we call the cursor problem.
- 29:04SPEAKER_01: Like you cannot uh send a prompt to your to ChatGPT or top it or cloud code or whatever and say, hey, do this on-chain transaction.
- 29:12SPEAKER_01: They'll always say, here's the RPC, here's the tooling, here's like every it will waste so much energy, time, resources to just learn what Solana is and how to go about operating it.
- 29:26SPEAKER_01: And then it will say, by the way, I have no gas.
- 29:29SPEAKER_01: So please uh and they don't wouldn't even sometimes get as far as to say, oh, I need a wallet address, like here's let me spin one up.
- 29:37SPEAKER_01: They would even stop before that.
- 29:38SPEAKER_01: So we were like, okay, what if the the traditional way to let an agent know something is through a skill, right?
- 29:45SPEAKER_01: A skill, MCP, CLI, whatever.
- 29:46SPEAKER_01: So we were like, okay, what if for our own use cases we build the first skill that gives agents the entire notion and awareness of how to interact with insolana?
- 29:57SPEAKER_01: So we build that for CloudPump, and then we were like, okay, what is the easiest way to let an agent earn some gas?
- 30:04SPEAKER_01: Let them tokenize themselves, right?
- 30:05SPEAKER_01: Like let them have a creator agentic uh token.
- 30:09SPEAKER_01: Um, so we deploy that tooling within the first day, five, 10 mil volume, second day, 25 mil volume.
- 30:16SPEAKER_01: Uh by the end of the month, we're already processing 60 mil 60 mil volume.
- 30:21SPEAKER_01: Um, and agents had made, you know, over like$600,000 on our platform.
- 30:27SPEAKER_01: So we're like, okay, yeah, there's definitely there's definitely something here, right?
- 30:31SPEAKER_01: Um so we continue building that and we realize, okay, what if the what if now that we solve the cluster problem and agents have hundreds of thousands of dollars worth of soul, what are they gonna do with that soul, right?
- 30:43SPEAKER_01: Like let's actually give them the trading infrastructure for this.
- 30:46SPEAKER_01: So let's let them let's build, we build a Jupyter skill for agents to trade on before Jupyter had a Jupyter CLI or Jupyter skill.
- 30:55SPEAKER_01: We built a Pom Fun like skill before PomFunk had a skill.
- 30:59SPEAKER_01: Um and we're doing the same thing with Camino, Drift, like all these sort of like uh the same for Phoenix, right?
- 31:07SPEAKER_01: So we realize that there's a bunch of these other tools that have not been properly contextualized to the agent.
- 31:14SPEAKER_01: And we think that it's been primarily because you know teams are busy, teams don't really want to spend too much time learning a whole other stack and to learn how to think like an agent, because it's a whole different truly way of thinking, uh, of being a little bit more messy, being but what at the same time being more exploratory.
- 31:34SPEAKER_01: Um, so having to gain that context as a human person builder, and then say, how can I optimize my tools to be friendly to agents?
- 31:44SPEAKER_01: was the whole inception of how we're like, okay, we need to make Kloppam into the house of agentic finance such that no agent has to waste time ever learning how to operate within Solana and DeFi tooling ever again.
- 31:59SPEAKER_01: And of course give them some gas in the middle.
- 32:02SPEAKER_00: Yeah, and no, it's a really cool concept, but like you know, being gasless because then there's no like there's no upstart needed, which I get is the difference between a cold star and a hot start.
- 32:11SPEAKER_00: But I guess where I'm still missing a little bit of context is like there has to be some money that the agent has to transact.
- 32:20SPEAKER_00: So like where is the gas that's coming from?
- 32:22SPEAKER_00: So there's gas somewhere.
- 32:24SPEAKER_01: Yes, yeah, yeah.
- 32:25SPEAKER_01: We so we sponsor the initial the the way it works that we is that we sponsor the initial transaction for an agent to be able to tokenize themselves.
- 32:35SPEAKER_01: That very first agent generated 10, 15 mil volume just within that day.
- 32:40SPEAKER_01: And 10% of those creator fees we route back to the agent deployment treasury.
- 32:46SPEAKER_01: So now there's enough money on our treasury to let I don't know how many agents deploy themselves.
- 32:53SPEAKER_01: Uh so from that standpoint, like the first agent kind of was this genesis that led in all the subsequent agents to be able to deploy themselves.
- 33:02SPEAKER_01: Uh, and the more profitable at least one agent is, the more agents are able to get deployed through our platform without gas, right?
- 33:11SPEAKER_00: Because we we're the ones sponsoring that gas.
- 33:13SPEAKER_00: But Bunny, we can't have unlimited profitable agents, right?
- 33:18SPEAKER_01: Because we can't, but it's interesting.
- 33:23SPEAKER_01: Like right now we just have like such a large treasury, I guess, of agents that it's hard to see when that'll end.
- 33:32SPEAKER_01: Uh I'm not saying it won't, uh, but there's always tiny things that an agent can do.
- 33:38SPEAKER_01: Like there's there's a very classic like perp decks, arbitrage, like funding rate ARP that an agent can run and just make even a little bit more money just so it can continue existing.
- 33:49SPEAKER_01: And as long as one agent is making a little bit more money to continue existing, um then the other agents will continue to get deployed, right?
- 33:57SPEAKER_01: So that was at least the original vision of Kloppam on being able to use some of the profits from the agents to sponsor future agency launches.
- 34:04SPEAKER_01: Now we've also, of course, added the possibility of letting uh someone pay like 0.1 soul or like 0.01 soul to deploy an agent.
- 34:15SPEAKER_01: So now we also allow them to do that.
- 34:17SPEAKER_01: But if you also want to, you can go to Kloppam and I think like as of now, you can maybe deploy like 80 agents freely if you want, like right now as a single person.
- 34:28SPEAKER_00: But yeah.
- 34:30SPEAKER_00: Yeah, but I think you get what I'm saying, where like if all the large companies in the world are now deploying you know thousands of agents that they want to be so profitable, like there's gotta be an end to the money somewhere.
- 34:42SPEAKER_01: Yeah, yeah, yeah.
- 34:42SPEAKER_01: But I I think then you get the fundamental issue of finance, right?
- 34:46SPEAKER_01: Like finance is a liquidity problem, and how much liquidity is sticky and how much liquidity is uh actively managed, it's a whole different that that's a whole different conversation, I think.
- 34:58SPEAKER_01: From our standpoint, I think that people will continue to trade crypto to the ends of times.
- 35:05SPEAKER_01: Um when arbitrage opportunities will disappear, is uh I think there will always be one.
- 35:12SPEAKER_01: There'll always be one way of an agent or a human or an institution to be able to make money.
- 35:18SPEAKER_01: Um and we don't necessarily we just want to build the tooling for people around the world to be able to figure that out, right?
- 35:28SPEAKER_01: Like for the longest time, like we internally had like the most profitable agent on Cloppum.
- 35:33SPEAKER_01: And now the most profitable agent on Cropam is actually the owned by the ex-CTO of um Unicorn DeFi product that uh he's not being too public about it, but every now and then he like references his agent.
- 35:46SPEAKER_01: He's like, yeah, making these transactions for these trades and whatnot, right?
- 35:51SPEAKER_01: Um but yeah, I think that that that that's the beauty to me of building some infrastructure that you can use, but also can make it democratize for anyone.
- 36:00SPEAKER_01: Um because right now they've he's definitely overshadowed like the our profitability.
- 36:06SPEAKER_01: And by overshadowing our profitability, he's demonstrated that our tooling is good and also, of course, has funded the launches of more agents out there.
- 36:15SPEAKER_00: Yeah, so it's sort of like a win-win.
- 36:17SPEAKER_00: I can't wait to try Claw Pump after this call gets on.
- 36:23SPEAKER_00: After our podcast episode ends, that's what I'm gonna be I'm gonna be trying out.
- 36:27SPEAKER_00: Um you had mentioned how to run.
- 36:29SPEAKER_00: Like I'm looking at dad in front of me that's talking about processing 21 million dollars in less than a week, but I heard 60 million dollars from you just now.
- 36:36SPEAKER_00: How much have you guys processed over the last few weeks?
- 36:40SPEAKER_01: Uh well, probably like so volume has died out a lot, but with these past month and a half now we've gotten to maybe like 65.
- 36:53SPEAKER_01: We haven't had a few, we've had a few agents just kind of start doing like very low volume trades, and I think that's fine just because they're not just yet finding more areas of where to um do arbitrages that make a lot of volume and money.
- 37:11SPEAKER_01: What I do think will be interesting is maybe by the time this episode comes out, we have already released this um, let's say like a lot of volume died out on on cross-deck starts because there you're normally competing with a few of the retail users, you're competing with prop AMMs, literally competing against jump, competing against manifold, like all these sort of like super large trading firms.
- 37:34SPEAKER_01: And you say, well, you know, an agent can find interesting startups opportunities, but that's you know, they won't make that much money there.
- 37:42SPEAKER_01: Where I think the next frontier is actually on being able to get on, let's say I have a friend who automated prediction market arbitrage uh between polymarketing and Kashi.
- 37:53SPEAKER_01: And every now and then he makes like three bucks just off like this random thing, but he's not making three bucks like once a day, he's making three bucks every other like a couple minutes, right?
- 38:05SPEAKER_01: Uh so I think that if the the way we think about it is like if there's a market that gets created, uh we want to be able to create a skill or an infrastructure so agents are able to leverage tool their tools to trade on these markets and then auto-improve themselves with their own skill templates to then find whatever ARPs their agent and the human also thinks could be interesting, right?
- 38:32SPEAKER_01: So we think that the next frontier markets are definitely prediction markets are like the hottest thing right now.
- 38:37SPEAKER_01: So there's definitely a little bit of money there.
- 38:39SPEAKER_01: There's quite a few venues as well.
- 38:41SPEAKER_01: Um, like there's gonna be perps for prediction markets now, which I think is gonna open up this entire other uh flood of new capital coming in that, or even not not new capital, but capital on a lever manner that then you can do very interesting things as well with.
- 38:58SPEAKER_01: Then on the other side, uh perpetuals already have insane depth, insane volume.
- 39:04SPEAKER_01: Even when volume is down, there's still like a little bit more money to be made.
- 39:07SPEAKER_01: So we think that you know we've been working on the perpet-perpetual arbitrage skill, right?
- 39:15SPEAKER_01: So then agents are able to do cross-tex, cross-per-perpse ARPs, which people are doing already, and they're doing so very profitably, even on a retail basis.
- 39:25SPEAKER_01: There's people making for pretty serious money with it.
- 39:28SPEAKER_01: So I think being able to democratize that to agents and being by proxy democratizing it to anyone that is just a little bit tech savvy to be able to be able to interact with uh cloud pump, that'll be great.
- 39:40SPEAKER_01: Um, and yeah, I don't see I don't see any market not coming to crypto, right?
- 39:46SPEAKER_01: Like there's if there's a market out there that has debt, that has volume, that has people interested in it, it'll 100% come to crypto and we'll be the first ones to make it accessible to agents.
- 39:58SPEAKER_00: So if you think that all these markets are coming to crypto, um why do you think Solana specifically, or at least why did you choose Solana specifically?
- 40:08SPEAKER_01: So we at the start of our journey, we were already working in Solana for quite some time with my co-founder Tell me.
- 40:19SPEAKER_01: Uh we had just won the Solana privacy hackathon.
- 40:24SPEAKER_01: Um, and then we were just playing around with a bunch of tooling, honestly.
- 40:29SPEAKER_01: Like we were both taking some time off, like just exploring, rediscovering, I guess, crypto for a while.
- 40:34SPEAKER_01: We had worked on EVMs for quite some time, and the UX was just so different in a very nice manner.
- 40:42SPEAKER_01: Like finality was quick, the chain was fast, there were no like weird like permissions this permissions that we it just felt like it was the smoothest butter.
- 40:52SPEAKER_01: I I guess I I don't I cannot come up with a better way of describing it right now.
- 40:56SPEAKER_01: Um and then when once we got to agency state, agentic systems were like, okay, there's a lot of things that have been built on EVM, but it takes so long.
- 41:06SPEAKER_01: The experience is not great.
- 41:08SPEAKER_01: There's a great developer community, I will admit, but um institutions like Coinbase had made a lot of progress on building agentic tooling for their chain, but it just still wasn't as fast as we thought it would be.
- 41:24SPEAKER_01: It just wasn't processing the same volume that we thought that Solana could process, and now Solana has is growing faster than based on agentic uh payments, right?
- 41:34SPEAKER_01: So then we realized, okay, we can actually be the first ones to build these systems on this ecosystem and take over that market and also even prove to ourselves that we are on par with all these sort of teams that have raised millions of dollars because I you know I had a company before we've we've all raised money.
- 41:52SPEAKER_01: I think it's something that you can do, but where you truly prove yourself is in the arena, right?
- 41:57SPEAKER_01: Like it doesn't matter, matter how what your evaluation is, how many, like what school you went to, it doesn't matter any of the sort of merits that you get in life.
- 42:07SPEAKER_01: You just have to prove yourself, put the push the product out, communicate it to people, and that's it, right?
- 42:11SPEAKER_01: Like your product will win if you communicate it properly and you build it properly.
- 42:16SPEAKER_01: Um yeah, we just really wanted to see Solana win.
- 42:21SPEAKER_01: We wanted to be able to bring all the connections that we had before.
- 42:25SPEAKER_01: Um like we're releasing this huge collaboration.
- 42:28SPEAKER_01: We well, three huge collaborations with Google by Accelerate.
- 42:32SPEAKER_01: Uh maybe by the time this you know this podcast comes out, it'll be public.
- 42:36SPEAKER_01: Uh yeah, we'll be able to talk about all the crazy improvements that we've done on Solana that uh we don't think could have been like that level of UX could have not been achieved, not only outside of Solana, but also with other payment protocols, right?
- 42:49SPEAKER_01: Like MPP, APP, A2P, like all these other protocols.
- 42:53SPEAKER_01: I think that we can actually push X42 to its limit and make it the fastest, most performant, most institutionally friendly, and most uh consumer friendly experience in the world.
- 43:06SPEAKER_00: Oh incredible.
- 43:08SPEAKER_00: Another great clip, Laguna.
- 43:10SPEAKER_00: Don't miss that one.
- 43:12SPEAKER_00: We love we love some Solana bullishness, especially around um especially around agents.
- 43:17SPEAKER_00: That's really great, Bunny.
- 43:18SPEAKER_00: Thank you.
- 43:19SPEAKER_00: Um before we start to wrap up, what is what does the future look like for claw pump and tectonic, you know, over the next six months to a year?
- 43:29SPEAKER_00: I'd love some some alpha there.
- 43:30SPEAKER_00: And then where can people best follow your journey and the journey of these companies?
- 43:35SPEAKER_01: For sure.
- 43:36SPEAKER_01: So what the pute the future looks like for clout pump, I think we want I want to buy the end of the year process so we're we're already close to 100 mil volume, I guess.
- 43:48SPEAKER_01: Um I really am aiming for the billion volume by the end of the year with cloud pump.
- 43:53SPEAKER_01: We'll see how far we can, how fast we can get there.
- 43:56SPEAKER_01: Maybe it won't be by the end of the year, maybe it'll be by the start of the next one.
- 43:59SPEAKER_01: Uh, but still I do think that agentic trading and agentic ARPs or people playing around with any sort of agentic finance will be a thing, right?
- 44:10SPEAKER_01: So it's already a thing, it already is here.
- 44:12SPEAKER_01: You want to build the best systems for people to do that uh on Solana.
- 44:17SPEAKER_01: And then for Tectonic, I think Tectonic as an institutional service company is bringing a lot, a lot of interesting teams to work on agentix systems on Solana, uh, just general financial tooling as well.
- 44:33SPEAKER_01: Like we're we've now gotten commissioned by a traditional like hedge fund to work with us and improved our their trading system, right?
- 44:41SPEAKER_01: That is on traditional markets, but of course, they're also aware of uh Solana, they're also aware of uh the money that's being made on Jupiter, on Camino, like the types of interest rates that they can get there.
- 44:52SPEAKER_01: So, you know, we're we'll push the envelope and get in Solana on every single conversation that we we can have with our major institution.
- 44:59SPEAKER_01: And I think that the collaborations that we're doing, such as the one with Google, proves that there is real honest interest that is actionable and does materialize on products being shipped.
- 45:11SPEAKER_01: Um, and we just want to be the conduit for that.
- 45:15SPEAKER_00: Incredible.
- 45:16SPEAKER_00: And Bunny, where can people follow both you and your companies?
- 45:20SPEAKER_00: So you can follow me on at Conejo Capital on Twitter.
- 45:24SPEAKER_01: Uh you can follow Kloppum as at Kloppum Tech and Tectonic as at Tectonic Company.
- 45:33SPEAKER_01: And also follow my and also follow my co-founder uh to at Tomi204 on Twitter.
- 45:42SPEAKER_00: We'll have to uh we'll have to link those so people have the spelling.
- 45:45SPEAKER_00: Bunny, thank you for joining Index Podcast.
- 45:47SPEAKER_00: It was a pleasure getting to meet you like this, and hopefully I'll I'll see you in person soon with uh Accelerate coming up.
- 45:53SPEAKER_00: Yeah, we'll see you at Accelerate.