Fireside Chat I Kartik Talwar and Stani Kulechov I Pragma Cannes 2025

[Music] But we're going to get started for the second half of the day. And without further ado, I'd like to bring on Stani from A to talk about all things DeFi. Please give him a round of applause.

Stani, this is a big week for you. So, um, I want to jump right in. Well, what does D5 2.0 mean to you and uh and why are we talking about D5 2.0?

Yeah, I think uh D5 2.0 for me means um there's basically a new environment in in decentraliz finance um to actually innovate and build. So obviously we started building DeFi uh already back in 2017 when when the coin when the term DeFi wasn't coined yet. Um and over the years obviously there has been some sort of um kind of like a headwind for uh decent interest finance and and and also uh blockchain and and crypto in general uh from like regulatory standpoint. So I think now we're entering and have been actually in a in a new type of an environment where actually um uh onchain innovation is something that is seen very um positive and there's a optimistic um atmosphere and I think this is the reason why we see a lot of new exciting announcements um coming even during this week and we see a lot of new innovation uh in the D5 space.

So for me the D5 2.0 know actually means that um it's the kind of like a new era of of being a DeFi developer and being able to innovate um and also kind of like a recreate ideas that don't necessarily exist at the moment uh in decentralized finance. So it's it's basically kind of a a new beginning to enter into the space as a developer.

That's awesome. Well, I mean the the other part that I think is interesting here is that you are working on a v4. Um it's been three years since the last version uh of uh a so um I'd love to know what's new and uh unique about V4.

Yeah. So one of the interesting things that happened over the years is that um AVA grew to be become um the biggest um D5 protocol uh at the moment and uh there is roughly 43 billion worth of u assets uh deposited into the protocol um uh providing yield against u different types of uh collaterals and there's different types of a market and markets and risk profiles and over time what we realized is that one of the benefits of of a and why it's growing so significantly is that it's a one entry point um a a um protocol essentially so or let's say the a product and what it essentially means is that um when all the users have one um kind of a market to enter the liquidity stacks into one particular um uh location and you have less of issues of uh liquidity fragmentation. So what we realized as well is that um in the future there's going to be a lot of more uh native crypto assets but down the line there's going to be a lot of also traditional assets coming on chain uh as RWAS and what we want to achieve is that we want to enable innovation um in a way where in a you can create different kinds of risk profiles uh without deploying new markets uh and segregating liquidity. So we created a completely new architecture uh based on uh liquidity hubs um and liquidity spokes and what it means is that a liquidity hub is where all the assets are actually stored and liquidity spokes are uh just a um uh borrow configurations or borrow strategies. And this means that um a spoke might have um these types of parameters of um uh what kind of assets could be listed, what's the loan to value of these assets are.

So how much you can borrow against um uh Ethereum or borrow against other types of uh assets. Um and then they're connected into the liquidity hub and liquidity hub is what gives you gives the spoke the the the credit line. So think about it as a um um kind of like a having a central bank and commercial bank. So central bank um creates a some sort of u risk profiling or capital allocation profiling and the commercial banks create these uh borrowing strategies. And what we really wanted to uh achieve is uh not only creating different types of risk profiling but enabling uh developers to actually innovate um and access to the office uh liquidity uh because we're meeting a lot of um D5 developers and there's probably a few of them uh in the house today and a common issue for a lot of D5 developers and founders is that um they build really interesting and amazing products but there's a u challenging and lack of of liquidity and typically it's solved by uh creating incentives or renting liquidity from um other kind of like a providers for for for additional um uh payments and that type of a model doesn't scale but if you innovate something u in a v4 by creating a spoke with some customization for your product um and if it's viable uh it can get access to directly from to the liquidity from the liquidity hub and that really seeds the uh innovation.

Did you kind of take inspiration from something when you kind of agreed on this model or how did this kind of naturally come to evolve?

Well, ma mainly the inspiration came from when you look at the existing a market. So whenever there's a new market um you have to somehow seed that liquidity and typically you have this chicken and egg uh problem. So what you see in DeFi quite frequently is that um a lot of um projects they use for example uh rewards uh or some sort of uh um liquidity kind of uh provisioning or or renting uh services. And what we wanted to achieve is that how do we actually um ensure that uh you have one central point of uh liquidity and and these markets could could create it at at any point at any configurations and seed it uh directly. So the the kind of awarding comes uh a bit more of uh the way the air airliners work.

So for example or any type of a hub and spoke uh system where you actually want to have concentration u but also like a segregation at the same time. I think the uh the central bank commercial bank is is one one kind of like a analogy to think about it.

No that's got it. Well, that's kind of one thing you already talked about, but I'm curious if you can give any other examples of spokes here and like what are kind of ways that we're already seeing people do hacky things around this in DeFi like can they replace this with a spoke model? Are there more ways to intuitively think about this?

Yeah, quite common um example is just thinking about the risk side. So uh for example a very popular uh strategy is to um um supply into a uh Etheas um as USDE use it as a collateral to borrow stables and leverage this uh position to maximize u uh yield and currently that asset is listed in the main of a market. Um and whenever there is a high interest rate it drives the the uh borrowing cost for all the users uh quite high. Um but with the AVA V4 hub and spoke model uh what actually happens is that we can create a separate um liquidity hub and and separate uh spoke particular tailored to this strategy. um meaning that the core liquidity hub can actually have a lower um yield but also lower risk profile while the Etheina uh liquidity hub and spoke can have a higher um cost of capital but also higher u yield and these uh spokes can be connected to any liquidity hub.

So for example, even in the lower riskier um uh liquidity hub can actually allocate part of the capital into uh the spoke if they want to have enhanced uh yield. So that's like a one example from like risk risk standpoint. The other example is um you can use for example um AMM LP shares as a collateral um by creating a uh a spoke for that or you can create a a spoke uh with for example uh safe or a custodian where the actual asset is stored u in the custodian or in nosis safe um while you're borrowing um funds from the protocol and this is an example requirement or uh kind of appetite that the institutional uh users are looking from like a compliance uh perspective.

Exactly. Because then then you can be on a third party custodian and still actually take benefits of all of this. That's actually a super cool idea. Um how does does governance change at all in V4?

Um yes and no in the sense that all these contracts can be governed by the avid but at the same time uh anyone can create uh liquidity hubs and spoke. So it allows um permissionless innovation. And the way I think about it is that you could bootstrap an idea um particularly from a u combination of of these uh both primitives um by innovating and if that idea actually gets traction you can request um access from the main liquidity hub that can be governed by the avid uh as an example. So there's different ways to to to think about it. Um yeah

how so so this is I guess I if that's the case so the thing I'm curious about is does that mean now you have to create a space for the spokes to govern themselves or would that still be managed by the entire set of contracts for for a DAO or do you because of the different risk profiles you kind of create different competing opinions on how some risk should be managed with uh how each spoke is set up. Yeah, the way I think about it is that there's two categories. There's one kind of like a of a curated sorry curated category of um liquidity hubs that users can supply liquidity and certain spokes are attached to that. But then there is completely permissionless uh liquidity hubs and spokes attached to that and the govern the governor is something where you can actually set your own governance uh mechanism. You can also set the avidal or um have a progressive governance towards um uh let's say to the aidal or whatever fits your uh use case.

So it kind of uh the new v4 architecture opens up the whole um like um development space and and just thinking of how to use the ache to whatever innovation um you might have. Um yeah, this is uh I'm still processing. So this is actually cool. Um the thing I guess what I'm trying to understand now is that the the time difference between V3 and V4 was 3 years and that's basically just a little bit longer than all the L2 stuff we've seen come up. Um and I'm kind of curious on like how do you think about what a crosschain world looks like with V4?

And does the spoke model help solve that a little bit? Does that make things easier? Does that make things complicated? Um what kind of does the multi-chain universe look like with V4?

Yeah, when we launched the AV V3, um we actually launched it first on L2s because we were betting that all the liquidity will move to L2s and obviously that didn't happen, right?

Yeah. But um the altus do exist and they they do have a significant value proposition specifically for transaction cost um or customization. We saw recently the the um um Robin Hood chain announcement as well where um actually institutions are also and and fintex are also thinking about running their own uh infrastructure for their own um use cases. In terms of AVA v4, we've we've always been uh agnostic to to the EVM environment and want to be wherever the users are and serve a particular use case. Um and the way to do that, for example, is that uh with this same architecture, we can have for example liquidity hubs across different networks and they can move actual liquidity between one hub to another uh from a network.

So it actually creates a really um efficient way to transfer liquidity between networks and you can do it in also way where um you segregate risk. So um you can give a um allocation of of liquidity from one network up to certain uh percentage. Let's say that hub number one on Ethereum gives only 20% um cap to be trans um transported to uh let's say to to base or arbitrum and and and and and this this is based just the same architecture where um you can actually um allocate from one hub to another hub um of of a liquidity with the same network or crosschain and I think this is going to be significantly interesting feature for anyone who is building any type of uh crosschain functionality, crosschain swaps, uh crosschain lending or or or fixed lending. There's a lot of opportunity there. So I guess that's a very super unique feature and it simplifies a lot of the common things that we see that end up adding brisk uh bridge risk.

Um and I think you kind of end up solving that a little bit because you're creating a line of credit now with a different chain on the same liquidity pool which can also have a spoke model. So it's uh this is if I understood it right this is like a pretty interesting creative outcome.

Yeah. and and um obviously you can do it in a in a different ways that you can actually uh create all almost like three credit lines to three different bridges

and kind of like a limit the exposures uh towards that and diversify.

So So it seems like you're kind of betting on this being a very common thing in the future and kind of coming back to the Deepi 2.0 theme. I'm I'm wondering like a few more years from now, how do you see DeFi evolve? This has been a pretty incredible year in terms of just stable coins becoming more popular, DeFi getting more popular again, a little bit more regulatory clarity. Um, do you have any predictions for what the next few years look like?

Yeah, I think what's going to be really interesting is just um using the existing architecture and use in infrastructure that we built in decentralized finance for native crypto assets and applying this to um to to also traditional assets that are coming on chain basically RWAS and reason why this is interesting is that there's a lot of assets in traditional finance that are extremely liquid um that could actually be brought on chain and one way we are actually doing This is with the uh horizon initiative that we have at a so horizon is a a market where you can use permission assets as a collateral. Um so for example Bidd Black Rocks Bidd um uh money market fund.

That's awesome.

Yeah. And and the idea there is that you can borrow permissionless uh stable coins. So it's a hybrid market and uh over the line um what we want to do down the line is that uh we're going to go higher on the risk curve um and introducing assets beyond money market funds um you know for example credits um and also adding uh assets that have more volatility. So ideally it's a market in where in the future you can borrow against money market funds uh credit funds uh private credits um and also uh ETFs um and and and um more kind of like uh equities types of uh assets as long as they have um offchain liquidity

because you kind of in a way offload compliance on the other end without actually making that a blocker and see that's what you're betting on

and the compliance happens um actually in the asset issuer level. So um if you are a customer of Black Rockck, you can access the Bidd then you can collateralize and and actually borrow against and um I think what's going to be really interesting um observation probably is that um how actually uh DeFi uh product like a can actually enhance the experience of colorization that is actually happening now in in traditional finance offering better uh competitive uh parameters. um and also adding more transparency and adding also um more flexibility. So for example, today uh in traditional finance you can't really access liquidity over weekends and I think that's just such an outdated

after 5:00 p.m. on

Yeah, exactly. or the next day and that's too slow for uh arbitrars that exist um in a shorter period of time.

Um maybe the last thing to uh to end this on is let's go back to the developers here. Uh V4 is extremely powerful. we just kind of learned all the cool things about it. Um, do you kind of have any ideas for what you would like to see from developers or like what do you think is like a message we can send or ideas for them to actually really explore uh the edges of what's possible with D4?

Yeah, I would like to see actually innovation um on the spoke level and and spoke is essentially a a borrowing strategy. Borrowing strategy can actually be connected to any type of uh other uh D5 use case. So it can be connected to a trading strategy or some sort of a um AMM uh product as well. Um and I think the innovation in this level is is um there's a lot that can actually be uh uh be done. I would like to also see some more innovative ways of uh um managing for example uh credit um using different types of assets as a collateral that maybe has been difficult to implement in the existing lending protocols um might be for example RWA assets some sort of a um NFTs as well uh depending what they represent um we're actually launching um

did you just casually mention that NFTs are possible to be a lending market or

no actually this makes it easier because Oh, I see. Yeah, cuz you can create

This is going to be extremely powerful if

Yeah. You can create a uh spoke um that accepts a certain type of NFT as a collateral and just plug it into a liquidity hub and get access to to capital.

UniS swap LP positions would be investing.

Exactly.

Exactly. And and typically why unis swap LP positions haven't been that uh successful is because they do have the chicken and egg problem. So someone needs to come and put the liquidity in and borrow and it's very hard to get the scale uh here. But in this new architecture you can create uh this type of a model and use case and and then plug into the um liquidity hubs and and I think that's really interesting. We're going to actually release a devet uh quite soon and publish the actual V4 codebase.

Um there's a lot of already um um documentation and content about how the architecture works and other additional features that we have in in in V4 that are extremely excited. Um but yeah, we'll be happy to help that anyone that actually are thinking of building on V4.

Um that's awesome. Well, that was my last question, but I think before we end this thing, I just want to say and recognize that it's refreshing and amazing to see that you've been still pushing the boundaries with like what's possible with a and like traditionally and even in crypto, you see people take less risks with changing things too much or adding too many things in terms of improving stuff. And the fact that every version is better even after you've gone bigger and bigger is to me absolutely incredible. So I want to thank you for uh pushing the DeFi ecosystem forward and uh it's awesome to see what's possible with V4.

Thank you. And I'm I'm also betting a lot on the developer community because um you know as from a labs we can innovate as as much as we can but I think it's up to you all actually to to push the innovation in the space forward. So thank you so much.

It's going to be a good rest of the year. So

I think so.

Thank you so much.

Thank you. [Music]