Welcome back to Block by Block. Today I'm thrilled to have Varun, better known as Gandalf the Brown on X. As the co-founder of Hibachi.xyz, it's a cutting edge decentralized perpetualist trading platform secured by zero knowledge proofs. Varun brings a wealth of experience from his previous roles and we'll dive into the product's innovative features and marketing strategies, community building, et cetera. So Varun, welcome. Thank you very much. Thank you for having me. Let's uh dive right into the product. Tell us what Hibachi is all about. um And uh especially for those that may not have heard about Hibachi. You know, absolutely. So, uh, like to use this the most boring way to describe Hibachi. It's a purpose exchange. So, uh, more interesting way it's a privacy first, perhaps exchange secured by zion knowledge proofs for the technical, nerds. this is basically a ZK roll-up, uh, focus on, uh, you know, private data, data availability. Right. So. So it's a sovereign ZK roll up on Celestia with a private DA. That's kind of the more technical way of thinking about it. ah really, think, mean, the way we think of positioning Hibachi is a consumer facing, it's an exchange before it's anything else. most, you know, anything we have been doing has been keeping that user in mind, which is like people who like to trade. I think, I mean, Peter, you requested a change on the platform. What about the open interest? Why is it not in the USD? And we got it done. So now you can actually see it in USD. So I think it's more user-driven feedback and to build a product. That's kind of how we've been doing it. So that's the general ah hibachi in a nutshell. Good deal. I first heard about Hibachi through my conversation with Nick White, who I've known for a long time. And he was sharing some of the of the things going on in the Celestia ecosystem. And he mentioned Hibachi is one of the ones that he's like really, really excited about. And so that's what led me to, you know, I had some curiosity, went to Hibachi and then, you know, created an account, started trading. I'm like, wait, this is pretty cool. It's actually like really fast. And I wasn't expecting that. And the fact that it was private um was also very, very interesting, especially given some of the issues with hyperliquid and James Wynn uh getting kind of liquidated. um And so was refreshing to know that no one could see my trades. And so let's talk about the platform. Because there's some quite unique kind of like uh It's on a unique stack. Maybe tell us about that, what that looks like. Oh, for sure. you hear me? My video kind of went black. uh Okay. Okay. All right. So, okay. So uh the stack is pretty straightforward, right? there's this, before I even talk about the stack, I think, let me start with the core principles of how we see the space. I think a general philosophy about building an application is compute off chain, verify on chain. Like that's the general philosophy. where anything that involves complex computation, doing a lot more uh delicate things. And like, don't use blockchain to do that. Just do it off chain. Right. So this A gives you a lot more uh flexibility in terms of iteration and allows you to build things that are more user driven and iterate fast. And B, it gives you the performance, especially for something like purpose exchange, where every millisecond is actually, you know, it matters. think having a central limit order book that is not constrained or bottlenecked by any kind of L1 gives you the most power. So, okay, let's run. So the first stack is AWS. That's kind of where we run the order book. So ah the second one is obviously we need to prove the integrity of the order book on chain because yes, we are running a central limit order book, but then why is this different from Binance? How do you know what's going on in the order book? and how the user balances are changing over time, they are in fact correctly reflected at the smart contract level where they deposit their funds. So this is why we need zero-knowledge proofs. And what ZK proofs are doing is they're proving the integrity of the order book, and then they are basically posting that state on chain. And they allow for anybody to verify that the proof was in fact uh computed correctly. So we're using succinct. to do this. we're using SP1 to compute the ZK proofs. uh And today we're posting these on Arbitrum. That's kind of where we accept all the USDT deposits, but we have a roadmap to add like other chains and then basically make it a more purpose uh architecture. And the last component piece is obviously the Celestia uh and the DA. So we're using Celestia for publishing the data and we... So the reason you need a DA is because we are posting a proof, great, but how do you know that there exists, like the proof does in fact point to the right data set, right? Like I could be posting proof of anything. So how do I know that, you know, it's the right data that the proof is pointing to? So you need that data sitting somewhere. And that's kind of where the Celestia DA comes into play. But now what you don't want, of what the traders don't want here is that they don't want their positions, that things that they've done to be exposed to anybody, right? So they don't want that to be public information. So what we do instead is we post an encrypted form of that information. So people know that there exists an encrypted data set that points to the right ah data that was used to compute the original ZK proof in the first place, but ah they don't need to really see what's in it. So that's kind of what the whole idea of this win upgrade in the private DA was. And that was always the first step uh in terms of uh getting to a foolproof solution. And there's going to be more subsequent upgrades coming up in the coming uh weeks and months that should uh harden the solution significantly. But that's the general stack of how the whole thing works and they come together. You mentioned something in there about Arbitrum. um Is that the first chain, I guess, that Hibachi is supporting and then you plan to go cross-chain? Yeah, at the moment we accept deposits in Arbitrum and users can deposit USDC on base, but then it gets converted to USDC. And uh in the subsequent upgrades, we'll basically natively support both USDC markets as well as just USDC deposits on base. And then there's also plans to add Mainnet and Solana as well, and maybe Bitcoin. So I think those are... things that are on the right path. Yeah. Do you, in terms of like latency and performance, do you anticipate any difference as you support Solana and other chains? ah Sorry, Peter, I you may have frozen for a second. You said, could you repeat the question? Yeah, you know, as you on the roadmap, as you continue to support other chains and fully kind of grow into a cross chain, uh props exchange, do you anticipate any latency or performance issues as you do that? ah not necessarily, no. And the reason is because, the core matching engine logic and the training aspect, it is not happening on-chain, it's happening off-chain. there's no uh constraints beyond what you're building internally. So we just need to make sure that the infra that we build is scalable over time and can sustain high loads. So as long as we can keep building on that. ah I don't expect any bottlenecks there, right? So, ah because the only thing we're doing is posting ZK proofs on chain, but I think pretty much every L1 or even L2 is capable of that kind of, because we're posting proofs every 15 minutes, so that should not be an issue. Let's dive into SP1 succinct. Maybe help us explain to the audience, from a trader's perspective, why is privacy important? What do they, I guess, prevent or avoid, or what benefits does it give them that their trades are private? Yeah, so generally speaking, mean, users don't want the transactions to be analyzed, studied, broadcasted by anyone on chain. So, mean, that's something that maybe some people wanted, but most people don't. And I'll give an example. If XYZ hedge fund has these proprietary strategies that they use to trade, what they don't want is a everyone to study their trading patterns and things like that to maybe copy trade, maybe even try to front run them, things like that. that's something that is not that that's not something that would like to happen. Right. So as a result, they would prefer that anything they're doing is like hidden from the public ah while doing the trade. And also once the trade is complete, you don't want that data to be public. So anyone can, wait a minute. This address coincides with jump trading and this is how Jump does all these trades. This is what they've done. So that is not something that they would be okay with. So that's one of the things. I mean, there's just different school thought, but some people think that transparency is required. I think from speaking to users, like they don't want that. So which is the reason for the whole push around a private DA as opposed to making that public to anybody. So that's one of the bigger reasons why you don't want. ah the data to be public or the users don't want the information to be public. But you could make a case that some people do want copy trading as a feature, right? Same thing, you've got a Bybit and Binance, you'd see like lots of copy trading ah as a feature. But the thing is, that point, you're giving the trader himself a choice. If you want to have like make money through copy trading, almost like running a hyper liquid vault, that is a very different ballgame than actually ah just making everyone's information public for anybody to see, right? That's something that is not ah something users would like to have, right? So that's one of the biggest reasons why we've been making this push here on the privacy. No, contrast that to, you if you were talking to a Web2 person that might be, um you know, does some options trading, they're probably used to privacy, right? um But then when they go into the Web3 or crypto space, then everything is like exposed. And so that's probably a big shock to them. And so this actually brings us closer to what TradFi does and expects. um um And entities, mean, you mentioned, you know, uh proprietary trading shops. I mean, it's, you you're right. They don't want to expose their strategies. so, it probably prevents because crypto is so open and transparent, it probably prevents many of them from entering the space, which makes me wonder um if any at all are even active on hyperliquids since everything's exposed. um But Um, it's interesting. I don't know who's actually active and who's not, but I, what I do know is that many firms that we've spoken to, uh, they did raise that as one of the things, know, before, before joining the platform is would their positions be public to everybody or not? Right. So that is, know, Yeah. Can you give us a sense of like how many... So each transaction is a proof or creates a proof. Is that correct? So the way it works is right now you do like, let's say 10,000 trades. And then what you do is you take that as a batch and then you compute AZK proof on that batch of trades, which results in, know. it sees the trace match correctly and then resulting in the user balance being changed or updated. then once that is done, that set is done, you basically prove that set on chain by computing that zk-proof, right? So it's done on a batch by batch basis. Got it. Is the batch size, is that arbitrary or is there kind of an optimal batch size? How does that work? Uh, generally arbitrary, right? So it's not like, we dig, but usually speaking, you know, we have like a set amount depending on, because we post proof every like 20 to 25 minutes. And then we just determine like, know, and the amount of transactions that come every 20 minutes is not like fixed, right? It could be different, but generally speaking, think, uh, yeah, at the moment, uh, we're able to prove all the transactions that happen every 20 minutes. Now, um in the, I suppose probably in the last year, year and a half since hyper liquids kind of become a big deal. I've noticed a lot of exchanges are like entering the perp space and, know, arguably, you know, perpetual trading is one of the kind of, it's probably the that and stable coins are probably the only two kind of crypto native products. that will be exported to TradFi that actually started in crypto and then now like TradFi is like very interested in it. um Tell us about like your view on the market with you know so many perp exchanges kind of coming onto the scene and how Hibachi differentiates. Uh, so today, I obviously, think, uh, uh, we are kind of competing with the same on chain defy users, all sorts. Right. So, uh, at the moment, I think the key that the two big pillars where we are differentiating, differentiating ourselves from other Dexys in particular is performance, speed and, uh, privacy. Right. So those are the three, uh, core pillars where we like, okay, we're saying is you're really fast. In fact, we're 40 times faster than hyperliquid. and we never put your transactions on a public ledger for anybody to see. So your positions remain hidden and you can trade at nearly 7 millisecond latency with no problem, especially if you're an API trader. Like maybe if you're UI trader, 200 milliseconds, 1 millisecond, you can't tell the difference to a human eye. But if you're connecting your uh APIs to trade, you will see the difference. And it's a big uh one. m That's where ah the two big differentiating factors lie today. uh But over time, I think the goal here is to build the exchange to a level of sophistication where it's on par with like Bybit or the Binances or the world where you can do things like spot markets, can do multi-collateral, you ah can basically have a full unified accounting system where you can uh use the entire portfolio to... uh take leverage positions across anything and then minimize the margin requirements and then maximize the leverage you can use. things like these, ah that turns exchange from like a DEX for some degents to basically going into full on sophisticated product that anybody can use is what would be the biggest roadmap or milestone in the next seven to eight months. And if you can get there, then that already puts us in one of the best DEX categories. So in DEX market shares increasing, And what you want to be is you want to be one of the top tier marketplaces where people go. So in general, I don't subscribe to the philosophy that only one exchange winner takes it all, especially not in the exchange game. It's going to be multiple winners. And it's going to be a function of some exchanges will do some things better than the others. And it's a matter of what users prefer. And at the end of the day, you want to be in that group of winners. That's, that's, so yeah, to say the least today, we're competing on speed and privacy as a core tenants, but over time, I think as things begin to converge, uh, you want to be, you know, a more liquid venue, better pricing, tighter spreads, more market types, things like that. And then obviously just building a brand that people love. I think, I think that's really where you want to be. Right. So that, that will over time have like dividends. Yeah, I agree. And I think on the retail trader side, I think one of the probably the driving, one of the characteristics that kind of drives trading behavior is liquidity and also availability of like the pairs that they want to trade on. um And that would be a different market than let's say, uh automated trading from, pinging the Hibachi API, for example. um How do you think about liquidity and how do you, you know, for, then select also kind of the pairs that you include in trading? Is the question, how do you determine what assets to list? Is that the question? so two questions there. How does that work with Hibachi, which assets to list? And then also, how do you determine, what does liquidity look like? how do you kind of, I think we had this exchange around now. I did a long on Doge and then um I noticed the liquidity was different than say hyperliquid. And so I was curious about how that works on the backend. how you determine liquidity. so hyperliquid has been on the block for nearly two years now. And they're obviously very liquid. They are hyperliquid, as the name suggests. uh And we've been live for six weeks. obviously, our liquidity is not nearly uh at the levels of hyperliquid today. So ah if you're comparing the liquidity levels today, hyperliquid is obviously going to beat us by margin. ah But that being said, ah Rome was not built in a day, like it takes. ah But we plan to get there. So, but that being said, in terms of what markets to list, it's purely community driven. So if there is a bunch of demand for a specific asset that they want to have a market for, uh we'll list it instantly. Like there is like no formal... diligent, uh there's no formal process where we are like basically spending a month figuring out what to add. It's very retail demand driven. Like if there is something that people are interested in trading, then we'll add it to the exchange and it's pretty quick. Like we can probably list something within like minutes. So. ah just in the Discord, is there a channel that says, you know, like asset requests or listing requests? there is something on discord and then it's also just in the general chat. people are like, okay, add this and everything uploading that then that's great because Luke is on discord channels. He's been talking to everybody. So, then yeah, every week we look at this, this, is the up and coming coin that we should list. Yeah. So going back to liquidity, just out of curiosity, does that work? I guess how does a purpose exchange get really, really deep liquidity? How does that happen? Oh, this is a flywheel, right? So initially, ah we bootstrap liquidity with some partner market makers who are on the platform. then uh once we're getting more take on demand, like this week, we've been averaging nearly 40 million a day in trading volume, which is quite good. So hopefully, uh it's going to keep going up. ah And if you have that kind of look, you're doing 100 million, 200 million day in trading volume, then you begin the flywheel. You have some liquidity to bootstrap and then takers came in, now they're trading more. And because it's more trading activity and take a flow, more makers are incentivized to put in more money to bring in more liquidity to support that flow because they're making money. And now all of a sudden there's more take more liquidity, which basically means better spreads and better rates for trading experience for everybody who's trading, which means more taker activity. And now more take our activity equals more market makers plug in and because there's more flow. Right. So then that flywheel begins and over time, you know, both parties are happy. You all of sudden have a super exchange, you know, a super liquid exchange. Interesting. Okay. I had always wondered about that. Um, cause I just didn't know the mechanics behind it. Yeah. of those things first. So you have some liquidity that attracts take a flow, the incentive us to take a flow early on. then once a take a flow expanse, more people want to bring in more liquidity because there is flow. it's like Airbnb is a great example. Airbnb, uh if there was no listings, no one would go. So someone has to first seed the listings. And then once you have some listings, some people book the houses, then people go, wait a minute, people are in fact booking houses on Airbnb. So let me create more listings, right? Then new people come start making, you know, adding more listings and then, and then you're like, oh wait, I have more choices now. So let me actually go on Airbnb again. And over time you have like a marketplace, right? So you can think of Airbnb listings as a liquidity, right? So, and, once you have like more liquidity, then you and me are the takers booking apartments and. places in Airbnb, right? So we are like the retail. Got it. um Let's talk about kind of marketing for traders. um I guess, how are you reaching the traders? You know, I came into Hibachi, you know, six weeks late, I think, you know, I'm probably still early, but still I feel a little late. But, you know. Yeah, that's good. That's good. But how are you reaching traders? um Like, what are some strategies you're taking to to do that now? can tell you a a few uh from now. I think one is obviously just aggressively talking to people on Twitter. think surprisingly CT is a great place to meet people. uh Commenting, talking to them, and then Discord. Again, just be in the trenches with your users. Talk to them, get the feedback, implement them. Obviously there's like some KOL strategies, you know, just talking to different, uh, probably, you know, like, for example, like we got in touch and then we're doing this podcast now. So it was pretty organic, the way we connected and so on. And then, you know, I've been doing the sort of thing with like many of the users who I'd say, you know, just talk to them and then see how we can get the word out and expand the distribution. And then we also massively working and tapping into the Celestia and succinct community because, because they've been our closest partners. And then they already have a broad distribution, which actually helps a lot in terms of getting the word out again, right? So, and then the whole thing has a compounding effect over time. So. Do you anticipate doing any formal co-marketing with Celestia and Sysynct? Maybe like a cross-pollination of communities or something like that? I suppose it's already happening, I'm not sure what do you mean by formal necessarily. um If uh you know to incentivize the the traders from the succinct community, know, maybe they get some kind of I don't know rebate or discount or something like that that's a good idea. We should do that. Thank you. So we thought about it, but I think we might actually implement something like that. That could actually work. And we can do it for both, even the TR holders, where we can actually make something for, like, Celestia token holders and succinct airdrops, happens. We can do something for those. So that could be nice. I ask because I do notice that there's a, I don't know what you call it inside of hibachi, but there's a uh timeframe where you could trade Tia and it's like, you get special rewards, something like that. So I guess you're already kind of doing it, right? Kind of. have not been like unique to Tia necessarily. It may just have been like a temporary market or like that we created. what he called it. Yeah, a temporary market. Got it. Yep. but like flash markets basically. yeah, but that being said, uh we could actually make something more formal and then just do that kind of thing. We haven't done it yet, but not, not actually, actually pretty good idea. So. uh Let's talk about Sparks. I thought that was really fun and kind of a dopamine hit, right? Like you open up the, you know, get Sparks and then you open up your magic box or, know, tell us about that. from... uh Actually, we did use that. That was another one that worked really well. uh Last week was kind of a nightmare. There were some issues that we had. So the demo was kind of on fire. uh hopefully, those issues have been addressed. Sorry, go ahead. Setting aside the issue. No, I guess, I guess maybe explain to the audience, like if as a trader you come in and then for every trade and kind of action you do, you get spark points. And then when you get like 10,000, you get a mystery box, which you open and then you get more spark points. Maybe tell us about that. Yeah, yeah. The idea is basically, it's like a casino box, right? Essentially. counting maintain that people like surprises. Right? So, so, I mean, anytime I was surprised, I received a surprise present, I was like, my gosh, this is actually kind of cool. Like when things are predictable, it's not as fun. But when you receive something, when you least expect it, or when you don't know, it's a forest gump. Like his life is a box of chocolates. It's kind of nice, you know, when you, so that was the whole general thought process behind Sparks as a program. Why don't we do, apply the same thing? incentives. Well, let people trade and once they unlock a milestone, they receive a mystery box, right? And then you don't really know what's in the mystery box, but then once you open it, they'll go, my gosh, I earned something new, right? And then today it's just points, but there's more creative things we can do with Sparks where there could be like more interesting gifts and things like that that come out of Sparks that is not limited to mystery box. And that could be really, really fun. Right? So that is something we're thinking about doing in the future. But I think that's the whole idea of Sparks is centered around creating mystery and surprising people with different things. And there are more interesting surprises in the books that haven't been unveiled yet, which is, as I still pretty early. There'll be more interesting things we could do. Well, I think you're tapping into the key kind of psychology aspect of traders, right? Like they, you know, it's mystery box. Like that's really fun, getting spark points. And I think like the next step is to have some kind of leaderboard to see, you know, where you stack against others. And I'm sure that's probably in the works too. um Okay. No, this is really, really cool. um So tell us what stage hibachi is in right now. I think it's, I mean, it's still in beta, is that correct? It is beta. Yeah, at the moment it's still like too early. a public beta and then um I guess what's what are the the next kind of stage after public beta? Uh, I think... After it's really, I think it's going into like a full production ready, you know, final boss, guess, because alpha is already done. Beta is right live right now. Maybe this is Sigma. I don't know if that is a thing in product launches, but I think uh we'll likely be in this version mode for about like another two to three months ah before we go on full on like, you know, production ready. But I think it's pretty close. All right, so. Yeah. And like, we are constantly iterating on the feedback that we've been receiving, but it's good though, because I'm like, okay, one thing we don't want to do is just spend too much time iterating and not shipping anything. So I'm like, okay, let's build. Yes. Things will break, but it's fine. That's kind of why the early people get rewarded more. Right. So, because if you guys are using it in this state right now, where things are kind of janky and still support us long-term, I think, I think there is, there is obviously, no, no, yeah, that's, that's kind of how you build. Thanks, Sideshow. Do you, you know, going back to kind of the audience and the target audience here, do you see a crossover or maybe like a very direct overlap of hyper liquid traders, um you know, potentially coming to Hibachi to trade and then, um and what are your thoughts on like, is it a different trader? I guess I don't know what the trader personas are. Like if the hibachi trader is different from a hyperliquid trader, for example, or a bi-bit trader, ah what do these personas look like in your mind? Uh, it's really hard to say that, right? Because in crypto, there's like a, what's the retail, what's institutional. It's like, like you, it's, I mean, uh, there could be a guy in his basement, but running like $50 million trades. Right. So would you call it institutional or retail, right? Because it's not an institution, but it's also a guy in the basement that otherwise would expect like a large firm to do. So, so it's kind of tricky to describe that persona. But I would say people who are using Bybit are like fairly sophisticated, know, pro traders, firms, but, you know, it's also been there for a while. So that's definitely like, you know, a more sophisticated audience, but hyperliquid did capture a bunch of that, right? So as well as the retail, but I would say hyperliquids retail volume, my prediction is that that's much, much higher. Right? So at least there's I mean, Bybit still doing a lot more volume, I would say Hyperliquid has a lot more retail focused trading coming in, especially for long tail of assets and so on. So, then Hibachi at the moment, we are essentially attracting more of the Hyperliquid type user profile ah before we go full on like Bybit uh adding like Hudson River trading, plugging in here. That's not the kind of users that we have today, but that whole institutional interest. Ah, we will have, you know, we're building towards that. You know, I first came into this interview thinking that Hyperliquid was your primary, Hibachi's primary competitor, but during the call you've mentioned Bybit multiple times and it, maybe there's something I'm missing, but it feels like that's kind of the, I guess the North Star is to compete directly. So that's kind of where you want to be. Right. and, yeah. And that's a huge, I mean, that's a, you know, these are formidable competitors that you're going up against. Yeah, no, no, for sure. I mean, it's going to be a tough battle, uphill battle, right? this, I mean, if it's easy, then there's no fun in it, right? So you want the challenge to be difficult. That's when it's most satisfying. So. Give us a sense how large, I'm not familiar with, know, uh perps trading on on ByBit or Binance, but like how large are these markets? Like, I guess what's the TAM? Uh, like really, really big, like, like more than $200 billion a day and volume just happening there between these three venues, Binance, Bybit and OKX alone. just in the, and this is like without like the whole institutional, uh, big players fully coming onto crypto, right? So that that's still, you know, It's still happening, if you're not, so which is why, mean, it feels like there's a lot, but it still kind of feels fairly early. Yeah, I was thinking small. I was thinking like, you know, I was thinking Hyperliquid was the main competitor, but you're looking at really, really big players. Yeah. I mean, hyperliquid is big. Don't get me wrong. ah it's like that meme where you have like the iceberg, but then if you look at the tip of the iceberg, So hyperliquid is, first of all, I do want hyperliquid to succeed. It's net positive for the industry. Like what they're doing is actually incredible. And I think it's a really, really talented team. So I think what they do, and hyperliquid even growing the DEX market share and taking away more users from these big exchanges or incumbents. is a net positive for any other DEX that has happened coming because it just proves that there is a market for DEXs. So I think by no means I see this as a threat. So I think that kind of competition is good and that's what pushes people in a way. And we want that kind of competition to be there. So because that's what's going to inspire us to do things better. So it's kind of good thing, actually. That's cool. Well, Varun, thanks for taking the time to speak with us about Hibachi. As a trader, I actually really like Hibachi. um It is snappy, it's fast, the UI is really, really just very friendly. I um was talking to a friend and I showed him what Hibachi looked like and he's not a trader at all and it was very intuitive for him. That's kind of a uh mis-test, I think. If like a non-trader looked at Hibachi, like would they know what to do? And he had no problem understanding what was happening. And so I think that's a really good litmus test. Awesome. Well, for the listeners, check out hibachi.xyz. uh They're in public beta. You can start trading today and follow Varun at Gandalf the Brown. Gandalf the Brown. And then where Brown is uh zero and not an O. Anyway, we'll put it all in the show notes. Awesome. Thanks, Varun.