Polynomial: Building a Derivatives Superchain

0:00 Welcome back to the permissionless podcast brought to you by the team at Guardian. And of course, once again, we're going deep with DeFi builders and DeFi teams, seeing what got them into DeFi, what kind of problems they face, what keeps them up at night, what they initially started building for, and what they're building towards today, and of course, where do they see DeFi going?

0:21 And today we have an alphapacked conversation with Gautham, a co-founder of Polynomial. And Polynomial is what they're calling a derivatives superchain. They've got their own L2 which they've obviously got their own pers and derivatives infrastructure deployed to. It's EVM based and it's a fork of Synthetix v3. And in this conversation we really got to dive deep into the considerations that the Polynomial team and you know most per teams are making in general with respect to the orderbook model versus the onchain settlement model with oracles and we got into the nitty-gritty of what the considerations are what the new products are on the market like Pyth Lazer and how the Polynomial team is thinking about this and where they're headed in the future and we'll be going much deeper on this in a second but of

1:10 My name is Owen and over three years ago, really almost four years at this point, I co-founded Guardian and we started on the premise that DeFi would envelop trillions of dollars of TPL from the traditional financial system. And so we set out to build the security audit solution that would allow teams to ship in that environment with trillions at risk.

1:33 We use two competing audit teams and a dedicated fuzzing engineer for invariant testing. Many projects get an audit and they still get hacked. The Guardian model is here to put an end to that. And today we've achieved the standard that is necessary to bring trillions in TVL into DeFi. And really at this point it's up to DeFi to catch up.

1:54 And of course without further ado, let us dive into this chat with Gautham from Polynomial. just to get right into things like what's your role at Polynomial? When did you start Polynomial? What what's the what's the day-to-day like? Yeah, like I'm I'm one of the co-founders of Polynomial. Oh, what do I do is I I do I do everything in a sense like I try to do everything, figure out how to do it and then get better people than myself to do that thing.

2:25 Yes, we started mostly like primarily little bit after DeFi summer because during that time me and my co-founders we had this small group chat of friends and we were just like farming everything trying out a lot of stuff and halfway through it we are like you know we can make better than better things at least like we thought we could make better things than the things and we were like okay let's just build something and then we try to build it.

2:51 We've been in the space from 2017 but at that time we were mostly in college at that time and it's more like you know we used to go to hackathons build up small small products and then afterwards we quit our jobs moved to like a house far off remote it was co you know you couldn't leave the house for like few months and then we were just sitting there and hacking and then we looked into crypto and then like one thing that made sense for us is like crypto is a primarily very good for finance at least much much better for every than anything else for finance and in every financial sector typically the derivative sector is the bigger sector than the financial sector like it's typically like 10 times bigger than the financial sector.

3:31 So at that time it just made sense that it's going to happen in crypto also it's already happening and so we set out to build products in derivatives and first we started by building on top of options slowly pivoted and is now building on purpose. So we launched our main product Polynomial trade a while I think it's been almost one and a half years on top of a chain which we made for derivatives and which is a per exchange and we have done around 5 billion in volume and it's been like we slowly like ramping it up but the core idea is the same thing like we think derivatives are going to be really big and in crypto like you typically need like specialized infrastructure for doing something like really great so that's how we

4:12 Are thinking that to build really good derivatives Yes, you need like very specialized infrastructure for that and it's kind of how we are doing at the moment. Yeah. And you say you you started with options. That's super interesting. I did not know that. Were you guys did you guys build your own like options smart contracts or were you building on top of another protocol?

4:30 What did that look like? we started off in a hackathon. we went to this ETH global hackathon and we made an options exchange on top of Uniswap V3 like a month or two after it initially launched and kind of made sense but it didn't work because options are really hard to pull off as an AMM at least because the it's like the volatility is very high the seller or the LP is always at a loss it's very hard to do risk management at these volatility levels that's kind of how we like moved out of options yeah and did you use like an oracle or anything for like the volatility index or No, no, no.

5:07 It was fairly simple. It was just like calling a put option and then you trade it and then but that was like in the hackathon and then when we actually built it, we launched a test net and we realized like everybody's making a loss. So then we actually built it for real because it from test net we understood that this is probably not a very very good idea.

5:26 I think like in a few years when crypto becomes le more stable I think it makes will make a lot more sense. Yeah. Yeah. Just the volatility index way too high. Yeah. And I imagine where you're doing it for like longtail assets on LP pools. I don't think you can do it right. Like I think sometimes like even on ether BTC voluntage it's like you're saying like in a year I don't know if it's going to go to double or half the price or like zero.

5:54 But yeah. So I guess was that the the transition from options to derivatives and and per was because basically you Yeah. Yeah. So we first build options and then we build like option ws and that is the first product that we made live. We built it on top of protocols called Lyra and Synthetix and then after a while we decided to make perps.

6:20 We first made on top of Synthetix. They were like our partner come before and then over while like building on top of Synthetix we realized that the primary problem is that we need to actually change things from the chain level itself and they were not really willing to do that. So we just made a chain and did it.

6:39 I see. Was was the main issue like latency with the trading what what was it? I think primarily the problem is that L2s are still very nent. they are not that great as we wanted to be right like in a comparative level it's still expensive before because plops are expensive right and so what can happen is that the fees for opening a position can just increase right like for us the thing was sometimes it used to take like $10 to just open a position so when you are having this $10 to open a position you price out a large amount of users and you also increase the liquidation price by a lot right suppose you had a liquidation price of $2,000 Just because a gas spiked your liquidation price is a little bit higher.

7:22 It can matter a lot for people with a large position size. So you do want a little bit of like specialtity and you can see it right like there are like platforms like Hyperliquid they are like not even using any chain to push it right or liked did the same. So an L2 is still a riskier option than an L1 where you don't have to post data on as but yes like it's still better than deploying on base or optimism which we did in the past like we deployed our thing initially first on optimism then we did deploy on base for a while but we did a private beta and it was way too expensive like so we just decided to like cut it.

8:02 Okay. Yeah. So it seems like mostly like gas and compute. gas compute then you can also customize it like I know it's like still like a very early thing but in the long term you can see a world where like every L2 will have its own custom layer it's not going to be like inside the sequencer it will be a little bit above on the top and you can see an early versions of this by flash blocks like how uni chain and base is happening so I think like this layer that is there it's going to be moreusto customizable to the usage, right?

8:37 Like it's more like if it's going to be like a derivative chain it will have something like for us it can be okay if there is a liquidation process it first before every other thing or if it is like you know we can make like you know these kind of set of rules which everybody agrees is best for the system and we can just process those thing really well right like for us one thing that we do a lot is updating price because we use an oracle based model for the part so we could like make it like happen much much frequently much much which decreases risk better for traders, better for LPS.

9:10 Definitely. And then I mean how do you how do you start to think about like compatibility, composability with other apps? Are you looking to bring them onto your chain like kind of like what Hyper Hyperliquid is doing? or are you guys the sole like builders on on your role or app chain? I mean like we would want to do it obviously and how we think about it is that primarily we are Oracle based and the primary problem I have we like any app has is liquidity right you want to build a new product but you don't have liquidity so what we did is we separated the liquidity part and the product so what you can do is the liquidity that we have is currently used just for the pulse but in the future if a

9:57 Options protocol comes or a prediction market comes they can still tap into the same liquidity. So what that means is if somebody comes and stakes in our liquidity layer currently just goes to the per but you can have like two per even right one for forex and one for crypto right and you can be like the LP can decide I want to provide liquidity to both at the same time and collect fees from both or just from one of them but the main thing is that if you look at crypto as an overall trend peaks of TVL happen in 2021 2022 like DeFi summer that kind of sign right And we never reached that level till now.

10:35 And though it sounds like like yeah it sounds like we made a lot of technological process progress over the last few years it hasn't like really materialized I would say and it's going to take a lot more time and I think like when that happens I think currently there is what maybe at that time 200 apps now I think there is 2,000 or maybe more apps right maybe 20,000 yeah maybe 20,000 everybody has their own version or something and all this liquidity just gets fragmented and fragmented away, right?

11:07 And half of us don't even know where our money is. That's why we have the DB banks and the zappers of the world to see which all chains we have money and all that kind of stuff. Increasingly useful. Yeah. Yeah. I guess through that lens really the position you want to be in is you want to be the holder of liquidity and then essentially people will come to you and they'll build the use cases off of that.

11:31 But you have the truly defensible thing which is the actual liquidity which is like that's your barrier to entry and that's the economic power that really all DeFi needs to build off of. So yeah that definitely makes sense and I guess how how do you go about attracting the liquidity? What's your basically what's your offer to the marketplace for aggregating liquidity?

11:53 I think what we do is primarily I would say like it's it's it's it's a bit complex of a problem right firstly like people trust you less because you are a new thing and then but one thing that we can offer is we can bring assets from mainet and then we can give an yield on top of it right for example we have SUSD and not a lot of apps on like L2s and all have like yield on top of SUSD even on L1 also right So may we are able to give like the trading fees it's more like 20%age on top of SUSD which on like a good time of the market can be go to like 30 to 40% right so it's like 50%age you can get yield and it's all like stacked

12:37 Up right so you can stack yield so I think that is like one thing and you can stack yield off of an asset from any chain and we do this because we make composibility a little bit easy we use socket and then we like you know have a mint and burn bridge on a side. So essentially what you can do is you can bring in an asset from base like you know some asset and you can have an asset from Arbitrum like we can just plug into our arms into any chain and then we can make it appear very seamless also right like when a person is depositing or trading in Polynomial from base they are not bridging into Polynomial and then going and trading it's more like you are just trading and it just

13:17 Happens under the hood right and it's like a good UX is what we are like trying to achieve and yeah that's kind of what we are thinking like essentially not having the apps like have like every time when there is a good asset I feel like they all grow in Ethereum mainet and they get the highest APY in Ethereum mainet and then after a while when the project has reached like peak growth they think about like going to arbitrum or base and then they finally deploy over there but then it's like they need to bring in liquidity pool they need to start giving rewards and it just like dies out over But what we can do is like offer like assets like this.

13:54 I think that is like a core thing that we can do and we are putting that under the liquidity layer and the same concept we are bringing now to trading. So currently to trade we are only allowing USDC but in a few days we will be launching this thing multiple collateral essentially imagine something like you have etheri which you're farming you can bring it into Polynomial chain and use it as collateral for trading right and lot of the times like perex they have wrap st or like most of the pers are in L2s right and most of these L2s like the liquidity is mostly in like either in USDC or right but you can we can like make it like you know you bring bring in like whatever token you're farming some new BTC or

14:36 Like some new ETH or even some new meme coin use it as collateral and then you can trade against it and that's kind of like how we are like thinking in compossibility terms in the short term. Yeah, that makes sense. So basically you're just limited to anything that's supported by what is it like a Pyth feed that you guys are using.

14:53 Yeah, we are using a Pyth feed. I think like still it's like something we want to improve on but yeah if something has a Pyth feed we can like get it up and we'll use it as a collateral we'll use it for staking but currently staking is essentially stable coin and stable coin like assets but in the trading side we can bring in any asset we'll soon like try to do it like much nicer and then we can have lot of things right like we can make it like a vault and then just like have basis trade right and currently like Ethena basis trade is take and shorting it why not do it with I don't know etheri or bugs BTC and then short it right that that would make more sense based on how

15:31 You do it right and we can offer them as vaults and all those things so a lot of unlock of what we were building is still pending and just coming and it's been a while so really on the UX side multi-chain sort of abstracting out what the chain is abstracting out your network in particular to the user and and then really how do we attract the traders so that we can get the volume to pay out the fees to the LPS.

15:56 It's really it to me it seems with these like purpose products it's all about the volume because if you capture a ton of volume then you can pay a ton of fees to the LPS. Is is that kind of way you guys see it as well? I think yeah I think at the end of the day like traders they just want like better prices and better fees.

16:14 So we are giving out like incentive programs in terms of OP and all but at the end of the day like traders what they really care about is primarily like speed I would say and speed has been something I would say like we've been like particularly not very good at till now and it's majorly because we use firstly an oracle and in an Oracle based burp you do something called a delayed order right so essentially somebody places an order and then it executes after a while and This delay is often like lifelong right.

16:46 So we are going to change it with like two big updates like one is like one is going to like at the end of the two updates I think like it's an audit at the moment I think like in two three months or like the first update will bring the time from what we have right now which is like five six seconds to like around two to three seconds and with the second update which is moving to a pit lower latency oracle I think we can go to around 250 to 300 millisecond I think like to be safe it would be like around half a half a half a second and that's little bit of latency everything combined then it would be a little bit comparable to Binance because if you're trading on Binance at

17:32 The moment you press a button you close a trade you press a button you open a trade you come on Dexus you are like you have a position open and you are closing and and if it's like a crash and the the seconds just just eats into your soul like you're like why is it not opening right why is it not closing it that just happens God forbid the the network gets like congested no because of the delayed order right like you're like well that and then also god forbid the the network gets on congested on top of that I I think the network congestion we don't have at the moment because we are siloed but the other one is still a problem yeah no definitely and yeah I did hear about that is that

18:11 The new I forget what it's called the new Pyth product that's like super low latency. What is it called? It's called bit with Pyth Lazer. That's right. What is the I actually I haven't looked into that. What's the technical change that allows them to go so low latency? It's more like you're doing offchain compute. You're little bit trading off on the decentralization I would say and then just since there are a lot of publishers you're just pushing it directly I think before a pitch chain consensus is done.

18:39 I think I am not 100% sure because we didn't start working on it fully yet. Yeah. No, I wonder if because I know the with Pit they have like their confidence intervals. I'm assuming they also have confidence intervals with Lazer. I wonder if they're just larger. No, I don't think so. It's it's more like currently since it's Solana which is a SVM thing, right?

18:59 So, I think it's more like you need to get the consensus fully done. I think that takes a little bit of time and the bigger time taking at the moment is like to fetch takes a lot. Currently, it's more like it takes like around 1 second to fetch the price. So it's like the price are fast but it's more like the infrar can be improved a lot.

19:18 I think that is with laser fixing making the infra like much much more. I mean that's pretty interesting because we see a lot of people moving to order books now and like specifically for the reason of like latency. Yeah. I mean I think mainly latency. So I think we are also moving to order books in a way because what we are going to do is like we are going to do a system where like when a person places an order we will have both an order book and an AMM and they will both compete for the order.

19:47 I see that's actually a really interesting idea because especially like in the Synthetix model you have this ambient liquidity that's over here and it's backing all of the open trades on the exchange. Does that mean that if somebody opens through an order book, are they still going to be backed by that ambient liquidity? Yeah, it will be similar.

20:07 Yes, it it's it's essentially like currently like what we have is a pool and then you have different you basically what we call as like a market like a per market. So inside the per market you can have many things and one of and sorry inside the per market you can do a simple standard order and then with that you can also have this like order book order which is essentially like a bunch of limit orders where you just match it right and that is kind of how we are thinking like if if the position is easily movable from the mm to the order book so you can start by just having someone just take the order book position and then just like filling it on day and taking a spread.

20:51 So essentially you can have you can solve the cold start problem for an order book because most of the time what happens in an order book is you launch an order book nobody's trying to nobody wants to take the other side and you end up having like market makers. I think hyperlink had had a unique approach of having a HLB wault.

21:10 I think this is also like one more way of like solving that particular problem. Yeah, because I assume you're also allowing them to to exit with either choice. So they don't need to find a like you open your position, you can also close it through the Pyth feed. So that's when you need the ambient liquidity to to be there.

21:28 How I think about it is like the AMM model is particularly very good for low liquidity markets, right? It's very good for like smaller markets. And when it goes to the top of the funnel markets, the ETH and the BTCs of the world, order books are significantly better. So what we think is think about a market cycle, right?

21:47 Like a market starts from spot at a certain point it becomes a per and for that the AMM is actually like much better than than order book because you actually have more like better slipage. you always know you have a position like you don't want to go into like a new token and open a long and then come back after a month trying to close it and there is no liquidity right and that's what happens in an order book sometimes and not right now but in the bare market like it used to happen a lot so that time you can just you know exit and trade via the order book via the mm and then after a while you just go and trade directly in the order book and then it just becomes more efficient

22:29 More sophisticated players can come and they are much more familiar with order book and that's kind of how we are thinking about it. Yeah. And then I assume liquidations would still go through the the stakers they would basically take on. Yeah. The liquidation but the liquidation fees is not that high. It's it's relatively like much much more optimized and it's a very small tip for both the person who liquidates and how the sticker is getting.

22:53 Yeah. No, that that absolutely makes sense. Yeah. So of course order books I mean really I don't know if we'll ever be able to beat the price execution price of order books latency but I mean with latency yeah that that Pyth feed seems like it's getting pretty close. Yeah, like we we are not taking a pure MM approach.

23:11 We are actually going very hybrid, right? So that's kind of how we are thinking about like it's going to be hybrid. It's going to be a combination of both. And with that like it's going to it's going to potentially like be a little bit better, right? Because there is no point in making the same product as Binance and hoping we will do better, right?

23:30 Because they actually are more centralized and they also have a lot lot lot of advantages. what we need to do is build something that doesn't exist at the moment. Yeah. And how would you expose that to the to the users that they like would you just complete keep it completely abstracted where it's like they they don't know if it's going to the order book or if it's going to you know onchain Pyth settlement.

23:53 How does that sort of play out in the UX? I think initially we would but over time it would become irrelevant and it will just get hidden and initially you want to show it to the user because they want to understand what is happening but over time I think what happens is people understand it and it just becomes second nature right and we also will get better at it and it's kind of how new systems get created right and I think a little bit of changes will be there in the UX it will be like maybe like a toggle or something and then just turn it on if you want to force Yeah, like advanced mode or something like this.

24:27 Yeah. Or it can be like a drop down in the market like do a market order then you have market order. Yeah, that's true. It could be like a different order type entirely as well. So just wondering were you guys always on Synthetix V3 or did you come from Synthetix V2 to V3 or or what was the transition like there?

24:43 We were on Synthetix V2 then we migrated to Synthetix V3 but we didn't really deploy the product on top of Synthetix V3 and then we made our own and then we so currently we are on it's it's a it's a bit similar to Synthetix V3 but there are like a lot of changes. Interesting. What particular did you guys I know there's there's a ton of like different things that you could do with Synthetix V3.

25:06 It's very like adaptable. You can plug in like arbitrary pools and arbitrary markets. What kind of adaptations did you guys make? I think for us like primary on the staking side we have lot lot more different assets like we bring it from mainet right I think that is one thing that we changed on the per side like it's more like the pit the laser thing the new order book thing those things are like very very different it's not there other version and a lot of some changes were like more on the more on the like the contract side it's it's more like how you on the what is it like the multiple collateral like there were X form changes like we we kind of like what we did was like we looked into like the

25:47 Problem and we were like okay we were building on top of Synthetix for a very long time and we picked up the things that we really like and we changed the few parts that we didn't really like agree with and like for example we didn't agree with having a SNX back stable coin for example and I think they are also like agreeing to the same thing but yes like and so like over there like that's what we did and then like little bit over time like they also did some things that we did and we also did some things they did.

26:18 I think at the moment I wouldn't say like they're like that similar of a product at the in like a few months at least. Yeah. Yeah. I mean well especially because they're mostly like purpose markets on like base and optimism and now I think arbitrum as well. Yeah. And you of course you guys have your own app chain right now and going for the the order book and and stuff like this.

26:39 So, and also the multi-chain. I don't think that like you guys are using socket. I don't I don't think that Synthetix has any kind of a integration like that. Yeah, we we we leverage a lot on the bridging side like we try to make things like much simpler and more focused on the UX I would say because in a technical sense I think like purps are decent but the thing is like you're not able to compare on a user experience level.

27:04 I think a lot of people do understand that nowadays mostly for D5 products. Yeah. What what kind of KPIs do you guys like measure to like compare to let's say you want to compare to like a Synthetix compared to like a GMX? What kind of metrics do you guys measure or put stock in? I think at the moment like we obviously everybody measures volume, right?

27:26 And that's the main metric, right? And then open interest, right? because volume is easily gameable at least some people do it. So open interest is something but at the moment like what we are focusing on is like the product that we have at the moment is more like an MVP. So we want to reach a stage where like it's it makes a lot more sense to the user like in a few months and what we are trying to do is like talk to the traders have chats with them ask them what they want.

27:52 We tried to build like strong relationship with like traders or stakers though we don't really know who they are because they ran on but we still like chat with them ask them like does this make sense all that kind of things. I think it's more what do you say that stage of the product cycle where you are more driven by user feedback than data.

28:10 Yeah, that's the most exciting part is like getting feedback of like a feature that very clearly people want and then you can ship it within like a couple days or a week. I mean obviously depending on how complex of a a change it is. If it's like UI stuff, ship it very fast, get that immediate feedback. That's a great feeling to have.

28:28 I built a few few software. I remember like few weeks back like somebody was saying like oh I want to trade on chart and then we just shipped it in like a week or something and now you can trade directly from the chart. You can now like you know place stop-loss take profit everything from the chart modify close your position like all those things directly from the chart which makes the experience of a trader like much much better.

28:48 I'm surprised we didn't do it before now that I use it regularly. Yeah. In retrospect. Yeah. One thing you mentioned is you've been building since like 2020 and on that level of like user feedback, what have you noticed people ask for now versus like back then and like the the evolution of what DeFi traders want? I think at at the end of the day like people are similar.

29:07 I would say it's I think people just want something that just makes money and the pro and the thing about crypto is the opportunity keeps on changing, right? The narrative keeps on changing. So your product needs to play well into the narrative. I think that is what people always want right like in terms of per it can be as simple as like list the new market title is trending so put tits right or like hyper right I think that is the main main thing that people really care about but if you really look at it like people are no longer interested in like trading from a metamas popup for popup but I feel like that is more like they are just being like sensitive to latency that is mostly on the side of like like

29:50 You know I make more money I lose money because it it took a lot more time to open right and I think those are like some things like people are but it's all coming back to like can you help me make money faster quicker better right and we see a world where people are paying 1%age fees on pump.fun telegram bots I think people mostly want like you know if I place a trade it should work should make money I think that's what they really really care about and that's been seen I think that that is the one thing that's been constant is the latency time to Open MetaMask has not changed.

30:23 Oh yeah, that is true. I wish there was a better solution. Yeah, I mean that's actually I mean like stuff like Hyperliquid how they have the like I don't know you sign a message and you enter trading mode and then you just and it just goes that's actually it's a really adept way to solve the latency problem above and beyond like an order book because it's literally changing the UX of like okay we didn't have to go all the way down into the literal inards of how we're matching a trade to fix this latency problem.

30:50 This was literally just a UX thing where we just remove MetaMask entirely and you save the user just as much time as changing from like a Pyth feed to an order book which is probably a lot easier to implement the the trading mode. I mean this is a session case right like they also implemented that. I think the one advantage they have is also like EDM like how we implemented session cases like with this account abstraction EIP and the account obstruction EIP is not very optimized for speed I would say it's it still takes a lot of time to like actually execute the order and that's why like 4337 don't have a lot of adoption also I would say and I think like that is one one other area like which you could like improve

31:39 But yes you can you could you can make the product UX a lot lot better and people have been doing it and also people at a certain point like people just wanted to trade perps maybe after the FTX crash on chain and they were willing to use any product I think the minimum standard of product they own is has increased a lot nowadays people won't just go and trade any random product now yeah the feature wars that's actually interesting I never thought about like how FTX probably drove a lot of onchain activity after they essentially rugged their their users.

32:11 I mean, that's how it happened. Everything started. I guess in some twisted way that was a little bit good for DeFi in that sense. I don't know. Yeah, no comment. Another thing you mentioned is like changing narratives in crypto. So, I guess aside from being a builder, as a user of crypto, like what narratives are you paying attention to right now?

32:31 Is there is there any other stuff on chain you're playing around with at the moment? Like I think memes are kind of overright like Solana meme narrative is kind of over. I think I can tell you like in a builder centric like what are some things I find interesting? I found fluid interesting where they combined decks and lending into a single protocol.

32:51 I think that is like a pretty clever way also find interesting. but I guess like in a way like you know these are like simple answers which everybody would say and there is nothing that particular other than these two that comes into my mind at the moment. Nice and I see you have the Mady PFP on is NFTTS like just Mady or is that something a vertical you're interested in?

33:17 Oh, I I I am just into I I don't really I never understood NFT during the NFD bull run to be to be really honest. I bought some NFTs, I made money, but I never knew how I made money or how I lost money and and then like me was something where I just liked it because of the people in it.

33:39 It was more of a community sort of thing than monetary or anything like that. So it's it's just just bought it and then I just kept it one day as a PFP and now I feel like it's very hard to change become like a part of you know your identity. 100%. Nice. Yeah. The lady once you put the lady ma mask on you don't go back.

34:00 You're in the you're in the cold. Yeah. From like a high level what what would you say is like the number one thing that's holding back Polynomial from like endless growth if there was one thing? That is a very good question. I I would say like it's multiple right like one is like core infrastructure like we need to improve it a lot more we need to like you know make the chain faster we want to make the block times like 200 250 millisecond saw that the traders feel like you know they are not in a really really worse of product than compared to like something like Binance right and I I feel like that is a good unlock another unlock which I feel like is the multiple collateral which I discussed about where you know

34:43 You can bring in any asset and trade against it. I think that would be like a really really cool thing and over like at least for this year our focus is majorly on the per but I think like afterwards we will focus on like having more protocols use our liquidity layer. I think that would be like a very interesting unlock over time but right now it doesn't make sense because the liquidity layer only has like few millions.

35:07 So it needs to be significantly bigger and then we will like start doing that. Yeah. So unlocking like the network effect of having other people come and build use cases for the liquidity and then I guess that kind of has like a compounding effect sort of a flywheel for you where people come and build that gives yield to the people who are staking and then you attract more stakers so people come and build more sort of goes from there.

35:32 Yeah. And also like simplifying staking is something that we were looking into like currently like you st you you make like you make like negative APY like staking has been like since the inception of product like like in a we actually measure like in 3 month period like it has not even gone once in a in a single day in a 3 month rolling period but the thing is like some people just come st and they just withdraw and the and might be like they made a loss right and it just like it's very sad for them and they lose trust or confidence in the product.

36:04 But we want to have like a simpler staking experience where someone can just buy a stable coin like product which just goes up in price and it's more like you kind of have a way where there will be like an insurance fund something like that and you make sure like the yields are like more fixed over time.

36:24 I think like in mainet you could do it by like maybe like a Pendle integration but like having your own chain comes up with you know its own compromises as well. No, definitely. And and I guess to get to that point where you're sort of like feature complete, ready to to have folks building on top of you, building integrations, and you know, you have the the UX that you want and all that stuff.

36:48 What What's the main like time ad for all of that stuff? Is it development? Is it is it the security process? What is it that like basically holds you guys down the most in terms of shipping? I think it's hard to make complex financial products like the thing that takes the most time is like to fix upon an idea right at least like that's how I code also like it takes a lot of time to like come and make an idea refined and then like be ready then you have a really solid idea to execute it just takes like a week maybe right and then or two weeks and then do audit also like two weeks or something that something of similar sense right you can ship it and you can make it relatively light

37:28 But to get the thing from a what is it like like some random thought to like a thing that would work is it takes a lot of time I think that is the thing that like takes the most amount of time right like it can be like oh let's just add oh what is it like let's just add auto book on top and it sounds like a simple idea it took us like a year to build it right like it takes a lot of time to and coding is not the thing that took the most of the time the six months seven months we just were just thinking and deciding While we were building other things, it happened on parallel, but that takes the most of the time.

38:03 Yeah, makes sense. Like sitting back and thinking about mechanism design, how do we want the incentives to be like? Is it are we going to do liquidations this way? Yeah. And trying to trying to foresee all the downsides before you actually code it up. yeah, I can definitely see how that's like a big bottleneck. So, in terms of like mechanism design, is it is it mostly you at Polynomial sort of thinking through these things?

38:26 Do you have folks do it with you? What what was your process look like there? Yeah, we have we have like around 13 people in our team and give or take like some people helping. yeah, like there is my co-founder. He also works on it. I also think about it some of the other engineers also kind of like we have this culture where like we kind of like do like small hackathons in the team every few months and then we build something we see how it works.

38:55 During that time we like typically brainstorm think and anyone can make something cool. Yeah, I think that's that's a great culture to have especially as you continue to scale like putting the ideation back on everybody else in the organization and like making sure they know that they can really affect the product in some way. I think that gives huge ownership and probably hopefully you see you know a good amount of ideas come out of folks just feeling that ownership and and wanting to contribute to the feature set that you guys have.

39:26 Yeah, that's definitely an interesting way to approach it. And then I guess in terms of like hiring, do you see any bottlenecks on hiring like any particular roles you guys are trying to hire out for? Yeah, that is really hard, right? I mean building a company is I mean I was seeing this tweet the other day like you remember CC was trying out I think DeFi or something and he was saying that he was finding it very difficult to figure out how to provide liquidity for an hour or something and then people were asking him like you know oh you don't really use onchain he's like no I don't really use any of this my I I am just good at hiring people and I think that is the truth like setting up a team

40:05 Hiring a people making sure there is a good culture. I think that part is the hardest part and it's it's it's always difficult and if I did nothing and just did that and did a really good job, I think like my job is done. Yeah. No, I I definitely agree. Ultimately, it's like how can you hire the people who are better at all the tasks that I'm currently doing and then you've upgraded every area of the business and then you're no longer needed, right?

40:31 So, yeah, definitely. I wish that was true, but there will be the business will have more things to do. Yeah, always inexplicably those things just seem to come up. But yeah, so I mean are there any particular roles that you're looking for right now? Right now we are looking to help someone in mostly like someone who can do mechanism design.

40:51 That would be like interesting. We're also looking for people who can do growth that is growth community like nontechnical side. We are a bit weak I would say because we are a highly engineering focused team. I think we would like to have like some really cool guy to come and take over the course if they want. Hey.

41:14 Yeah. Well, me too, dude. Me as well. But if if there's is there anywhere we like let's say if there's anybody watching who would be interested in like applying for any of these roles, is there anywhere we can send them? just DM me on Twitter. That would be the easiest to you can send an email to jobs.fi.

41:35 You can DM me and say like listen to this. I would at least take a look. Yeah, for sure. Well, I mean hopefully hopefully you can you can get a superstar just based on somebody happening to to take a listen. But yeah, dude, I I totally agree. I think hiring's especially in such a new field where like you're you're literally not going to find anybody with more than like five years of experience in DeFi.

41:59 I I think that is more than that. I feel like the problem is like in crypto like people jump from one thing to another very fast. Yeah. And the thing is like when you are in complex topics it's very hard to be really good at it in a year or something. And if you're not really good at it, you're not going to solve it really well.

42:17 And that is a problem which I feel like unless unless you are like some expert from beforeand and like you already are like really good at it. But I think like people need to have a little bit longer term point of view. I would say I think that would be like better because crypto is already has very high learning rate.

42:37 Right. Typically like people used to jump from one company to another because you are so bored you know everything. But in crypto I would I would say like that's not the case like you like two days like two months the narrative is something else. People are on to something new. Somebody came with a new algorithm which makes things much better, right?

42:57 And it just keeps on moving. I think that part is super interesting about crypto and I think like it would be good to have like some people on our team. Yeah. Bottom for the long term. I guess it's all about incentive structure when it comes to that. So I mean how do you how do you think about incentive structure for incentivizing folks to to stay for you know four or five years?

43:18 I we just do standard stuff. It's crypto like you make something good you you know you will be sent to like there is no question about it right like it's like you you are seeing literal billionaires created out of one year or two years of effort yeah it doesn't happen in any other industry I am good at I know of right like like you look at AI like there are billionaires been made but Sam Alman and Ilia has been working on it like 14 years or something and everybody's like researcher PhD like that kind of like level, right?

43:52 But on the other hand, like crypto like people are like it's it's a good field where like you have you have the pump.fun where you are you are like doing what around a billion in revenue for an app that is made less than a year. I think like that kind of growth I don't think it's possible outside and it's I think it's worth it.

44:14 If if you want to do something in your life, crypto is the best place to do something. And that's I strongly believe it. That's why I'm building in crypto. Yeah. The best like I would say I don't want to say risk adjusted return because it is a very like risky like you don't know what's going to happen in the space in in any in any given year.

44:33 But yeah, I mean if you're going to take like 10 shots in any industry, I feel like in crypto you're going to get the most amount of those shots be like a home run than anywhere else. And and also like the problems are so obvious, right? It is not like people don't know the problems. Most of the time you are trying to find product market fit like chewing glass in darkness, right?

44:54 And here in crypto everybody's telling like bro this is broken. Everybody knows it's broken. It's not like nobody knows it's not broken. It is more like does anybody actually want to step up and fix it? Yeah. And and who can like pull together the team of people who have like the right experience to fix it? because that's like you said, it's hard to like attract those people in the first place, find the right ones, and then incentivize them to stay for a long period of time.

45:18 But I think there's we have an abundance of problems in the crypto space. That's what we've what we've established. Yes. Yeah. And then I mean really finally like what is your overall like strategy to to attract like the right people? Have you tried anything in particularly? I mean, I know we've tried a few things that I'm I'm happy to share that, you know, like work out for us, but anything in particular you guys have tried?

45:43 I think we have to refer. Yeah, I will say one thing we we started doing for like we've been trying to recruit like lead researchers who are like the best in the world. and that was very hard for us because I mean, well, first of all, there's just not a ton of them, right? And they're they're barely like they're very competed over from, you know, top security firms and stuff like this.

46:03 What we tried is we put out like a just a big like bounty, referral bounty. We put out like $20,000 as like a referral bounty and basically it just got a lot of like attention and from that it got the right people's eyes on it and they actually reach out to you directly. So we never actually ended up having to pay out the bounty.

46:22 But just putting it on there shows like hey we we mean business. we want this. And we got the the attention and and we were able to get like a few of like truly truly exceptional like worldclass researchers off off the back of that. So that that that's just what worked for us. I don't know if if you want to try it out, feel free to to try it.

46:45 But yeah, that's what we've been playing around with. I think that's a good idea. But yeah, dude. I mean, really, I think we're we're coming to the to the end of our time together. But is there is there anything else that you want to cover? Anything relating to like I don't know business and DeFi problems that you see market outlook anything like that?

47:04 Nothing in particular. I'm just curious what do you think about some interesting things like I see a lot of people tinkering with AI nowadays but it's more like the agents of the past where glorified chats. Yeah. I I I can't tell if it's like I mean like you said there's just so many of them and the thing is I don't even know if any of them are like legit.

47:27 Like I guess the play is that we'll find like some superior UX for interacting with blockchain. I'm still not totally convinced on that, but I don't know. We'll have to see. I guess it really comes down to like who's got the users at the end of the day. So if they've got the users, then they can direct all that liquidity.

47:43 Then I guess they've proven it out. But I haven't really seen any AI app that's got a ton of users. One interesting thing I find is like people are just making these Twitter boards where they are public, right? So you don't really have the users. So it's Yeah. But yeah, that that's that's a that's an interesting space.

48:02 I have been like just looking at for a for like fun because I I used to do a little bit of AI in my previous job. So I was like super super curious about that. Otherwise like do you see any new trends in DeFi in particular? Yeah, let's see new trends in DeFi. So the the way I see it there are different areas of DeFi like the obviously like different core primitives.

48:27 I think in the derivative space everybody is they're going to either they're exploring order books like we discussed they're exploring multi-chain and we're working with a a couple derivative teams right now who are working on both of those things and then I would say in general multi-chain and layer zero in particular they seem to be really cornering their market and bringing a lot of assets to multi-chain at least with like the type of work that we've done like we've just for one reason or another we've done a ton of layer zero work in the last like two three months on a number of different things working with like crosschain NFTts like crosschain I won't say I won't say what asset is crosschain because I don't know if it's it's public yet but the biggest NFTs

49:10 In the world will be will be crosschain soon I can say that yeah and crosschain like per like I mean obviously you guys already have it but more people are exploring that I think they're sort of catching up to to the feature parody that you guys have on that front but obviously you guys are working other things ahead of that.

49:29 So I think crosschain is a good narrative. I mean there are like lot of decisions you can make also like crosschain BS like even though we have we make it like you know people bring the capital and trade you could have a system where like you have the capital in that chain itself and trade the trade as in the risk engine isn't there itself like you can do something like that that is an angle we didn't try but yes I think there are a lot of things like that it's interesting that like you know are there any you know interop tokens at the moment it's not there right like the the crosschain tokens is that like a really big thing.

50:01 Not really. I mean, there is zero, which I guess the incentive for that is just like you can use zero to pay for crosschain transfers, but like nobody does it. You always use ether. I feel like the problem here is the liquidity. Like you can't really spend like insanely large amount of money on every single chain to get liquidity, right?

50:19 So like people are just people just try to just have a hub for liquidity in a one particular chain and then just like leave it at that, right? Yeah. I think all the value unlock probably in the next couple of years is and what we're seeing people building towards is just on the UX side in terms of how do we abstract away gas, how do we abstract away the chain and then how do we also get you better execution price and like like how do we make it easier for you to make money?

50:44 Like the best DeFi app is like you you push a button and like you get yield immediately. So I guess we're we're like building closer and closer to to that experience. Yeah, that's what we're seeing for a lot of the people, a lot of the teams that we that we work closely with. I mean, like we've done like half of it.

50:59 I don't know. I think like at the end it's also about the market opportunity. I would say the narrative leads to Yeah. chasing their narrative as well. I think that is I would say the narrative right now in DeFi is crosschain in my opinion. But yeah, I appreciate you hopping on and and having this combo. I think covered a lot of things in terms of like how you're attacking UX and what you see the outlook is in terms of getting the best experience for traders, attracting the most liquidity, using that as a liquidity layer for other folks to build on top of.

51:31 But yeah, it seems pretty bullish and of course if anybody wants to go use Polynomial you can check out polynomial.fi and trade or provide liquidity and get paid some of that yield. Thanks. Thanks for having me.