Building on Zircuit A Hackathon Introduction I Jan Gorzny

Uh thank you everyone for coming to uh building on Zirket a hackathon introduction. And so I'm just going to give you a quick introduction about what it means to be talking about Zirket at this hackathon. Um I'm going to cover three things. Uh what a great hackathon project is at large, what uh Zirket resources you might want or need, and a little bit about our prizes. I think that's why uh most people will listen to this talk.

So first what is a great project? A great project in my mind and I think the mind of the judges at circuit which happens to be me this time uh generally solves a problem. Uh hackathon problems uh serve a lot of purposes right? Some people do it to train themselves, some of it to change the world, some of it just to win the prizes. But I think at the end of the day anyone that really solves a problem is the one that's really valuable.

Now those that definition of problem is very very variable. Um, you know, it doesn't have to be one that addresses a market uh need, but it certainly could be, right? Real projects and real products were launched at hackathons like this one. Yours could be the next um it could also just be something you want to do to overcome some technical challenge. Whether that's you personally struggling to to trade things on different networks or an interface or integration between two systems that currently is giving you a lot of friction in your day-to-day development or user experience.

And finally, the last problem might be you and that you're inexperienced and you want to learn something about it. I think either way, any of these problems is worth solving. And uh I think the best way to think about if it's a good enough problem to really tackle to especially get prizes is determine if it's something that you're going to be thinking about when you leave this building. It's very easy to build something over the next two or three days, however long this is, and on Sunday or three days, come to me and say, "Hey, I've got this really cool idea. I think I built something."

And then move on. I think the best projects are something that remains in your mind or my mind for days or weeks or months or years to come. And I think as long as you do that and you understand what problem you're trying to solve, you're going to be successful even if you don't get a prize. But I hope you get a prize. With all of that in mind, don't overengineer your solution.

There's a lot of buzzwords in tech and in this space, uh, don't throw things in just for the sake of doing it. Don't try to target multiple sponsors just for the sake of trying to win many prizes. No, have a well- definfined goal. Don't overengineer it and stick to one use case and doing it really, really well. Showing that you're competent in tackling one feature is a lot better for me as a judge to see that you can build all the other features than if you just try to sort of make a a stew of of different features using different things.

And when you do go to solve that feature, think really carefully about how you're going to do that. Uh first you you may want to, you know, look at various tech stack options and be attracted to a lot of them. You know, a lot of things like to use AI, but do you really need AI? Right? There's there's situations where, you know, a regular expression can do the job or you can just hardcode some stuff.

Um, when you go to use things like blockchain or other things, you know, justify it. This is a web 3 uh hackathon. Probably you're going to use a blockchain. Certainly, if you're trying to win a zirk prize, you're going to use a blockchain. Uh, that being said, you know, if you're doing something that can be replaced with a static uh database because you don't have trust issues with people and decentralization doesn't matter, think about that, right?

Maybe that's still a good project. Maybe you still want to hack it, but maybe you're you're going to be less excited about um the prizes. But I think even if you did that, um you'll have an easier time going forward with this product or project when it's done, whether or not you get private. And then validate and listen. Validate your idea and listen to people.

Um talk to other people. Get, you know, get a feeling that actually this is a problem point for not just you. There's been a lot of times where people will say, "I really don't know how to do this. There's no solution out there." And they're like, "Wait a minute.

There's there's a project over there. or they're doing exactly what it sounds like you need. Go check them out. And it's really upsetting because they wanted to build this really exciting thing, but I just said, "Someone solve this problem." As a former academic, I also know that, you know, it's really rough when you have an idea and you think it's interesting to solve and then someone says, "Here's the result, and it's done."

Um, that is really frustrating. But you also want to make sure that if that's not the case and your project is worth solving, that you have uh some people who like it, right? It doesn't it doesn't necessarily need to be everyone. Uh depending on the type of problem you're solving, some are very technical, some are very personal. Um not everyone will say that's a problem I have or that's a problem I need solved.

But some should, right? It shouldn't just be you in an echo chamber because I think um that might be too local, might still be worth doing, but it might be very hard to to think about going forward. And I think good projects should not be forgotten. Along those lines, listen to feedback. Some people will say that sucks.

Some people are wrong. Uh some people will say that sucks and those people are right. Uh don't you know listen to everyone without thinking about that information critically. You know there will be some people who give you valid criticisms or say you could do this simpler or you could do this better. And sometimes you just can't do it over the next 3 days because the simpler solution is harder to engineer if that makes sense.

But maybe you know there's a simpler solution down the road but over the next two days you have to get the hard solution. Or vice versa. Maybe you know you're doing something hard and there really is an easy solution that you can use to accelerate this. So listen, don't necessarily take everyone's feedback into account. If you are getting feedback from a judge and you care about that prize, maybe listen to them a little bit more.

But uh certainly, you know, be a collaborative builder because ultimately if you're building in this space that's largely permissionless and largely open to everyone, um understanding other people's needs might be very valuable, right? You might build for one particular wallet and find that your biggest users really love a different wallet. And that is like something you can change, for example. That's pretty easy, but you wouldn't necessarily know if you didn't talk to people. Admittedly though, time is limited.

You only have a couple of days. Um, don't necessarily go around all day just talking to people about ideas. But, you know, turn to your hacker next year, get a coffee break, go to the water cooler, that type of thing, and say, "Hey, I'm building this. What do you think?" and they might say it's good, they might say it's bad, and you have to reflect critically on that, but I think it's valuable.

Um, that being said, if someone does talk to you about an idea, don't steal their idea. I don't think that's the goal of this, and I'm not trying to encourage that type of behavior. Be reasonable humans. And finally, I think the best projects at the hackathon are going to be built on circuit. Circuits of zero knowledge roll up on Ethereum.

We do a lot of cool stuff. The basic stuff is that we are performant and cheap. We have 2C block times. transactions are very cheap, less than a cent in most cases. And we're totally EVM compatible.

So everything you want to build on Ethereum, you can build on Zirket for a lot less uh money and certainly a lot cheaper than on the mainet itself. And you can always use our test net. So you know, it'll cost you nothing. And you can get test net tokens via um various faucets including one through the ETH global hackathon website. But beyond those unique features which you might have from other places, we are unique in that we have sequencer level security.

So we have systems built into our sequencer that will filter out malicious transactions. So if you're building a DeFi protocol or you're otherwise concerned about your users's funds or accuracy of your your code because you don't want to hacks, um we have a system that's built in to quarantine transactions that are going to do damage to your app or your users. Hacks should not exist on Circuit. That's our purpose-built mission. We don't want that.

So if you're building something with large financial incentives or you're building something that's risk adverse, consider Zirket for this reason. This is something no one else will offer you. This is something that only we have. It's very very effective. And if you have any feedback on that or questions or you want to learn more, I could give 20 minutes of talking just about that topic, but I can't do it right now because I have only so much time and I want to talk about the prizes still.

Um so do all of this and you know utilize the unique functionality if you can. There's actually no SDK and there's nothing you need to do. So don't be frightened by this AI integration. It just works out of the box. But as I said, great pri projects are also ones that win prizes.

So let's talk about the prizes, which is might be why you're here. But before we do that, I'll give you a little bit of a hint about what circuit documentation and resources you need. So for documentation, your go-to place will be docs.ircuit.com.

It should be up to date. If it's not, because we have been developing and we also are rebranding. So if you go to circuit.com and it doesn't look like my Korean cats, please don't be scared. But if you see uh anything on the documentation that is out of date or needs to be updated.

For example, here I see ZKVM that has changed. So that's already gone. That's how old this slide is. Um let me know and we'll fix it. Part of us being here is to get users on board and to get you excited about web 3.

Uh another part is to get feedback for myself just like I want to give you feedback when I'm judging your pro. Otherwise, to connect to the network, it should be as simple as changing the endpoint. Your wallets might have it already in. If not, put the endpoint there and put the block explorer and chain ids. It should work out of the box.

Any issues here? I definitely want to hear about it. Please let me know. But it is literally these four or five symbols that should need to change between, you know, circuit and any other EVM compatible network that you're used to. So, it's not hard and should be very easy.

We're really proud that we also built out a custom block explorer. While it might not be exactly like Etherscan, we think it's better in a lot of ways. It was built unique for that sequencer level security. We had to put that somewhere. And it's a little bit uh built for all twos that use zero knowledge proofs.

So you might see things like where proofs are being posted and things like this. But nonetheless, this is your main resource. And if you have, you know, any concerns about that also feedback, but it should work and have all the core functionality uh that you'd expect on anything else. Bridging if you don't get um tokens on circuit through the faucet directly, certainly on you can bridge um or from mainet. We also have other bridges, but we have a cononical bridge that you can always use.

Uh, and this is the URL, bridge.circuit.com. Really easy to use. Faucets I already touched on.

ETH Global has one. I think it's probably the best place. If you need more, come talk to me. I'll be at a booth. I may be AFK a little bit.

Uh, but if I'm there and you say, "Hey, I need more uh uh Sapoleia ETH or Circuit ETH or Circuit Testnet ETH, let me know. If you want to go live on mainet, you want a little bit of ETH to do that." I might be able to help you there. I don't necessarily want to say that on record, but come see me if you really think mainet's the place to be. All right, so purpose of this particular hack, uh, sorry, this particular talk is the prizes.

A lot of you are probably here to see what the judges, in this case me, wants from uh, you guys. So, there's three categories and I'll talk about it, but generally speaking, these requirements exist everywhere on all three of those and I think they're pretty reasonable. If they're not, let me know. They're not necessarily strictly hard. If you miss one or two things, they're probably not going to be disqualified.

Uh, but what we'd like to see is some smart contracts. We're building a web 3 thing. Uh, those should be deployed on Zirket or Zirket testnet and they should be verified via our block explorer. This should be a straightforward process. Some code obviously uh you're building something with code.

Let's see the code. It should be open source. And then we should have a read me describing what you're doing. And I think it's very straightforward. I just want what you've done, why you integrated Zirk or how you integrated Zirket.

uh who you are, what's your background, was it your first time building something or are you experts? Uh how do I actually test it to make sure the code works? If something doesn't work and you expect it to work later, say that I understand that there's only a couple of hours to hack. Like it's a very short period of time. I'm a human.

Treat me, you know, as a human and you can give me commands and say, I would have built this, but I can't, etc., etc. Also, come show me your project at the end of the week. But um anyways, that's another story. uh if you have time or you want to pitch this uh you can also do a video demo or a slide deck.

One thing that um isn't strictly in scope for this talk is we have a grants program at Zirket. So if things go well here and you want to get even more money uh come talk to us about that. It's a different system but uh having a a slide deck or the feedback from the act can help you a lot there and then deploy the stuff. So what do we have this time around? We have $10,000 in prizes.

Apologies if you checked earlier and is only 8,500 is in fact 10K and it's split amongst uh five winners and three categories. The first one is best project. I think this is actually incorrect. I believe it's supposed to say best DeFi project. Uh check the official website for the prize, but I'm pretty sure it's best def.

The good news is this description is otherwise more or less exactly the same. I just need some application that works in this case with some DeFi. So put some stable coins in there. It's pretty easy to do. Have a well- definfined target audience.

Is it DeFi that does you know some particular type of yields? It does some type of loans is it an integration with unis swap before you know have some ideas um and then do it very well and that's totally subjective admittedly but um we'd like to see new ideas if possible right if you have a game changer that's huge but if you can't do that because new ideas are very very hard uh something that you've improved upon makes sense as long as you can tell me what you improved upon. If you're just forking something, this is not the best prize category, but there is a prize category for that. This one is accurate. I apologize for the last slide.

Best use of account abstraction. We were really proud to have uh support for Perra as soon as Ethereum, more or less. I think we're a couple days behind, maybe a week behind. Um, nonetheless, we put EIP702 support there, which is probably now part for the course on most L2s. But, you know, I think it's an underutilized area.

I think people talked about account abstraction for years and years before these EIPs got included and now it's here and no one's building EIP stuff. So I would like to see some EIP 7702 um integrations on the chain because I think that validates the community effort if not Zirket's efforts to build it out. Um the downside here is that depending on who you use as a pay master and some 7702 integrations, Zirket might not be listed. If you use some wallet that has a preferred network, maybe it's not there, but there are command line instructions um listed on the bounty description or surprise description, sorry, that you can check out. And we have seen people develop really successful AA based um approaches in past hackathons despite this limitation.

Now, I really hope that's not a limitation. And so my last category for which there's only one win, one winner, um is to overcome all those limitations. Circuit has been around for a year and a half. We're really proud of all the work that we've done. Um, you know, I think we're fairly popular, but if you look at L2B, we're not number one.

Let's get to number one. So that every time that anyone has any tool that says deploy on a network, Zirket is a default option. I know there are tools out there for which Zirket is not a default option. And I want that to change. So I want you to put Zirkuit into that.

If you find a tool that says uh open a PR so that the endpoints of Zirket is there, that's as simple as we need here. Um, it can be five lines. just put the endpoints and the block explorer and that could be the winner. Um, if you want to do something more involved like somehow connect something in some way that requires more code also good. But the purpose of this uh prize is really to make sure that if there's a tool out there connecting to a blockchain that you have the power to modify because it's open source make sure zircuit's there so that we can have circuit everywhere.

I think that's incredibly necessary for you. So in this case I really don't need novelty. This also includes forks. If you go on our chain and your D5 protocol is open source, um your favorite one, sorry, it's open source and it exists on another chain but it does not exist on circuit fork it deploy it that counts. Um this is the category for which you know novelty is not there.

If you want to improve on it, I think you should try for the the better uh D5 projects. But really this is about making sure that if there's any libraries, codes or tools that you would normally use on another blockchain but Zurka doesn't have, we want to fix that. and I know what some of them are and I'm working on some of them, but I know you guys deal with way more tools than I do. So, help me help you. And that's it.

Uh, so if you have any questions, I'll be at the booth. I have a couple engineers here. Um, documentation link is there. Feel free to reach out personally on email at this. Otherwise, um, I'll jump really quickly to the front so that if you ever have any concerns, hit me up on Telegram over the weekend.

Just say, "Hey, I'm at the hackathon. I have an issue." And either I'll come by the booth to see you if I'm not there. And if I can do that or I'll give you the answer on Telegram. If not, I'll point you to someone on our Discord and we always have people on call to make sure the network's fine.

Someone will answer a question very quickly for you. Uh so thank you very much and I hope you have a great hackathon and I hope you win some prizes.

We're done. [applause] Any questions?

No.

Okay. If anyone has any questions,

I mean it it could uh so it's meant to stop hacks that would also have been pushed through a pay master or using an account distraction based approach. But the AI is sort of oblivious to the type of transaction rather the outcome of the transaction what it look is what it looks for. Okay. Thank you.

Hi. Hi. um that sequencer layer thing you mentioned

sequencer level security.

Yeah, it sounds very interesting. Can can you give me just a little bit more detail?

Yeah. Yeah. So, we built in an AI that's been trained on years and many blockchains like billions of transactions over the last 6 eight years. I don't know how long it is now. Um where we had all the transactions that made your favorite WC.

News hack happen and we said those ones are really bad. We don't want those on our chain because I don't know what a hack is. You probably have a different definition than I do. Well, we can all agree if it makes it to wreck news, someone lost a lot of money and a lot of people are really upset. So, we took all these transactions, six billions of them, said the ones that made those uh news entries are really bad, let's stop having hacks like those.

And uh we trained trained a model to make sure that it'll catch outcomes like that and it'll stop those types of hacks. And we know it's effective actually because we can have the same um system which we built out with partners looking at other blockchains and they detect those hacks in real time. So we can validate that it worked. So we would have been able to stop the balancer hack on our network or at least a significant portion of it because the AI actually flagged that as a bad network or a bad transaction, but balancer wasn't on our network. So I couldn't stop it and I couldn't do anything about it.

Um it's it's it's that effective. It's 99.99% accurate. If you get falsely flagged as a hack, so you're honest, but it's not a hack. You'll be released from quarantine typically within 15 to 20 minutes and your your thing will work.

So maybe for the most high frequency trade that's a problem, but honestly stuff that's pretty basic generally doesn't get flagged. I think somewhere, you know, we've had maybe 13 million transactions in the last year on our network. We've had maybe less than 500 false positives. So that rate of error is like really really low. And you know, we'll always improve it and all of that, but uh for now it's like a strict benefit for almost everyone.

Nice. Nice. Nice. Thank you. Thank you.

Any other questions? [applause]