Scaling Cardano: How Peras and Leios Will Transform User Experience

Hinzugefügt: 2026-05-09

[00:00:00]All right, everyone. This interview is going to be really exciting. We're going to learn a lot about uh some core development in the Kadano ecosystem. And I think Paris is probably one of the things that a lot of people are really interested in in the future of Kadano development. And I have the guys here from Tweig. I have Alex and Chris join me on this episode to talk through what they do at Twig, what they're developing for the core Kadano infrastructure, and we're going to learn a lot of things here. Uh, but gentlemen, welcome to the podcast. Thank you, Peter. So, I'd like to start off with a general introduction of what Tweak is, what you guys have been working on, and we'll start with that so everyone gets a good grounding of where we're at.

[00:00:50]Good morning, afternoon everybody. Thank you Peter for having us. Thank you for making the time. I'm here having a twig hat. Well, it says modus with twig and I'm here to walk you through our 26 28th proposals for the Cardano core infrastructure. I do have other hats.

[00:01:05]I'm sitting in the budget committee helping steer and build a budget processor that's more welcoming to everybody. I'm also one of the member of Cardano Cura constitutional committee.

[00:01:18]This is just to tell you I'm not just here for to present the infrastructure projects. We have the skin in the game much more than just doing the dev work.

[00:01:28]Before I get into specifics, I want to frame why we're asking for what we are asking because this isn't a shopping list of 17 separate projects. It's one delivery pipeline with 17 moving parts.

[00:01:39]We have been in business since 2013. We have been continuously engaged with IO on Cardano since January 2018 nearly eight years. Our team led the consensus and ledger work implemented or bar genesis contributed to design of orburas.

[00:01:56]Uh we have been contributing to layos to multiple projects smart contract audits etc. Since this 22 when twig joined modus that gave a broader digital transformation backbone designer strategist AI specialists while keeping the engineering discipline the twig is known for I say that not to brag but because what we are proposing depends on deep familiarity with the core node. You don't want someone learning Cardano internals on a consensus layer operate.

[00:02:27]Cardano next growth phase has two engines. Paras for faster finality and Laos for higher throughput. Paras takes the settlement from roughly 12 minutes down to 2 to 5 minutes. Laos dramatically increases transactions per second. Together they unlock kind of responsive user experience that defy L2s and partner chains like Midnight need to complete compete seriously.

[00:02:53]But this is where our proposal lives.

[00:02:56]Getting Paras and Laos to exist is not the same as getting them to mainet. And getting them to mainet is not the same as getting to the ecosystem to adopt and trust them. A lot of 25's treasury spending went into designing those these upgrades.

[00:03:11]Our next proposal is about delivery, hardening and activation and why this bundle is nonmodular. The work packages fall into three for reinforcing categories. We have protocol delivery paras v1 to mainet and paras v2 for resilience history expiry to keep node operation affordable as throughput grows. Safety scaffolding as a second part where we have conformance testing mutation testing hoarding node suit for network observability and the hard fork incident meool bridger as a safety net. And the third thing developer and operator tooling the Pluto script reexeutor canonical ledger state for mitral and amaro integration and maintenance of tools the ecosystem already depends on you can't ship pairs safely without these testing scaffolding. You can't safely increase throughput without history expire or your price small operators out of running nodes. These pieces are designed to lend together.

[00:04:17]And when we were building this, we were thinking about the vision and KPIs. This map this maps clearly to pillar one infrastructure and research excellence and pillar five ecosystem sustainability and resilience. The KPIs we are moving are the ones the community has asked us to focus on monthly uptime transactions per day TVL ALT full nodes clients throughput capacity and annual protocol revenue. The ask and the close.

[00:04:46]The total ask is roughly 39.8 million ADA about 9.9 million USD at the 25% conversion spread over two years. That's 17 work packages staffed by a team that has been on this codebase since before many of the alt clients existed. We have separated what we control code releases benchmarks from what depends on governance like hard fork activation.

[00:05:11]you seal that split coverage in every work package and Alex and I are happy to walk through any of them in whatever depth makes sense. Okay, there's a lot to go through here and I'm just going to go back and ask some specific questions and uh help clarify some things for a lot of the listeners. I know my audience is all at different levels. They hear orus um they get confused. They don't know what's going on here. Laos, we got Paris. We we need to break this down a little bit. Um first off, Paris is the finality side of things to get that confirmation that a transaction has happened on chain faster than what it is at the moment. Can we can we get a bit of an explanation about that? Where where is Kadano at the moment? Where will Paris take it? And how does it compare to something like Bitcoin? Can I just give a real world analogy before Alex takes on? When I heard first first about Paris, it was a it was a big uh discussion with researchers and now where we are building it. How to explain it to somebody who is not very deep into the blockchain there. So think about difference between a wire transfer sending bank to bank or doing a card swipe in a shop on a point of sale. Wire transfer will settle eventually. You will send the money and the money disappears from your account and x amount of days that appear somewhere else and you know you sent it. You trust your bank. You trust your client's bank that you'll ever receive it. But when you swipe card at a point of sale, it's gives you a near instant confidence.

[00:06:46]Okay, the machine says paid, you get the receipt, you get your stuff. And this is actually what we're doing. Parases Cardano's card swipe moment. We are lowering the time from 12 minutes to five.

[00:07:01]That would be easier to understand.

[00:07:03]>> Gotcha. Yeah, I I like that um explanation. makes uh sense and gives it give gives people a visual idea about as well. Alec, do you have anything else to add to that?

[00:07:12]>> I I want to add and emphasize a few things. Cordana is a probabilistic network. So basically whatever you put on the chain eventually you should get an understanding that it will be applied everywhere and it will not be rolled back. And as for all probabilistic chain that gives you additional nice guarantees you always have this probability that something will be changed and currently in order to get from the wire transaction to a swipe cards. It's an additional protocols that are built on top of the network and what Paris gives Paris gives a guarantees as much stronger guarantees from the network itself. So basically it's a thing that you shouldn't reimplement in your own protocols. You shouldn't build your own guarantees. You get all the guarantees from the network.

[00:08:01]>> Understood. Thank thank you for that uh explanation as well. Now, there were some other things that you mentioned, and this one is something that's close to me because I'm a small state pool operator myself, and I find it kind of concerning sometimes when we have layoffs and we have >> the we have layoffs and Paris coming on board and the potential cost of operating and running a node might become very high. What is the modeling?

[00:08:28]I know you're working on this to to make sure it's as cost effective as possible, but what does that look like at the moment?

[00:08:35]>> While we're waiting for Peter to unsync, we're already proposing Paris V2. And as we are putting V1 to the mainet, V1 V1, V1 would be like we built the house and we're moving in. But V2 is let's weatherproof it. There's there there is a storm coming. We live we live in a in a zone where there are storms. So we're going to weather proof it against wind against water. We know when we're avoiding the cooldown periods where finality temporarily pauses if a voting round fails. Why? Because if an attacker with about 25% of stake could force Paris into cool down repeatedly, we are implementing a pre-agreement algorithm that reduces how often that happens.

[00:09:20]Essentially giving an additional layer of security. I'll elaborate on that uh a bit. Currently, Paris v1 is a thing that has everything that is required to be on the mainet. Still, there are known limitations of the protocol that exist.

[00:09:39]As Chris has already mentioned, it's two things. It's a cool down period that we may get in and it's how we can get out of the cooldown period. And there are some knowns algorithms, but some of them are understandable at this point and can be implemented, but it's much bigger than being on the network. And some of them require additional research and at this point it's important that we can start do can lead uh some of that uh right now.

[00:10:14]Basically when Paris comes to the main mainet we should already thinking about the next steps as otherwise it will take a year or two years for going through the old cycle of the research and then testing and then development and getting to the network. And this is important that this work can't be just postponed and we will think that we will do it someday else later because otherwise it will really take too much time to get Paris fully done.

[00:10:47]>> It's really good to get an explanation of how Paris is working and all those moving parts. One of the things that you mentioned a little bit earlier that I was concerned about when reading into Paris and Laos and how that was all all going to work was the cost for small staple operators. You mentioned that and I am a small staple operator and it kind of scares me when I realize that the ledger will get larger. We have to, you know, store more data. The resources might get bigger there as well. What type of work are you doing there to try and minimize those costs? You say that you are, but I'm feeling the costs are going to go up for us as well. So what's the future here? Look look look.

[00:11:25]>> So we propose one additional feature that expect to reduce the cost for the stake poolool operators and it's a history expiry project. This project began in the 2025 uh intersect propo funding proposal but at that point we understood that initial proposal was not good enough. it didn't get all the agreements within the community and we rethought that approach and we noticed that this project is especially good for the in light of layers and pairs coming to mainet. Uh the idea is that we still keep the header chain but for the blocks we can reduce can drop parts of the history and then the question will be what should I do in order to get the old blocks that are reduced and we had another project that we just finished in the 2025 proposal that is genesis sync accelerator it's a creation of so-called archival node basically a node that can keep all the blocks and all the data on CDN that is much cheaper that doesn't require you to run all the resources and use all the resources that are needed to run the network but is able to cheaply reply with data. So this project that we have just finished will be a big stepstone to reduce the operational costs. There is a lot of work to do and it's not something where you that you can just plug in and simply say that you've done. We need to understand that all the current protocols still can operate. How the network will operate if there will be many like shallow so-called shallow nodes that do not have all the history and we will put all the amount of research and uh implementation that is required there. We expect that this thing is not bounded to the hard fork times. So it can come to the life and to the network once it's done but we plan to have it done roughly at the same time as we have we will have layers and paras to get a wider adoption understanding and this way smalls will be able to run the servers and be happy and do not lose all the money on operation cost. Okay, there's a lot of moving parts here and so we we've got some new terminology as well that we need to get our head around. So shallow nodes and archive nodes. Um when when you started talking about the archive nodes and pruning off that history, it kind of gave me Salana vibes and I was thinking, oh no, you're going to move this to a database, but it's actually going to be what? A separate node network. So the archival node will be a separate node not uh it will be it will look like an normal node on the network following all the protocols. It has nothing special. It doesn't have any additional database or something and it can be run by any provider who wants to run this node.

[00:14:34]>> Gotcha. Okay. So um it might be something that I run in conjunction with my main state pool but just holds all that data. So have the one instance that will have larger file system compared to my um other nodes that will operate the >> and the main thing here is that you can use not a big drive but you can use existing CDN networks where the drive may be much cheaper than keeping your own big drive.

[00:15:03]>> Very cool. Okay. All right. That's uh that will definitely bring down costs.

[00:15:07]Okay. I love the engineering engineering here and all the uh the thought that's gone through this. Now, with all this, there are a lot of moving parts. There are a lot of things that you have to piece together to make this all work.

[00:15:19]And the amount of work and coordination that you guys have to do to make this happen is just mindboggling to me. Like, uh, you're trying to work with another team that is doing the layer side of things. You've got all these historic factors that you need to integrate in and then all this new stuff that is just finished researched and you're trying to implement and then integrate that into a fully operating and running chain. How many pieces here are actually involved to get this up and running and how do you manage this? I'm just dumbfounded by you know this complexity here and I I just want to get an understanding of the amount of engineering that is required and amount of effort cuz I think that really does justify the costs that are coming up with what you guys are trying to do. I can say on one part as we have been doing this for some time and actually the community is not that large the the developer community we could probably all fit in a very big conference room and currently we have in my slack I think we have 20 different channels connected to various companies where we collaborate and it's um how to say on some parts we are the leaders in some projects we are the workers and that goes for our partners in the ecosystem. We we work as a big hive. I would say that would be the best way that's the way to do it. We are segmented as separate companies but we're all global. I think the big motivation is the love for Cardano.

[00:16:51]That's that overcomes all the issues we have especially with speed and how fast we things are moving and that helps a lot having open doors. I mean we are able to talk to everybody first of all we talk to we're open to talk to we collaborate with everyone and that makes it much easier just be like a open book.

[00:17:10]>> Okay. It's uh kind of nice that you do have all those channels and you know everyone in the space otherwise yeah it would make things very hard. The other thing that you mentioned was the timeline of things here and you said something along the lines of 2 years. Is is that for the amount of money that you're requesting now to be paid out over 2 years? Does that also mean the development is going to take 2 years to coordinate and launch all of this? Our engineering is very mindful of their time. First of all, when we were doing this, we scraped everything that was didn't have impact or users already. Our engineers want to build stuff that don't have reason to exist. And our planning is very very good. We are usually inside few weeks of our initial planning to the delivery with all of the PTO's, vacations, holidays, whatever come up.

[00:18:02]We we we always add buffer because you cannot estimate 12 18 24 months with high precision. For some things we deliver faster. For some things we need extra time because unexpected things change. The ecosystem moves.

[00:18:19]No, some things can move. So we this is our target. We could deliver sooner and that's good for the ecosystem but we unlikely deliver after because that's how we do that's how we plan. If we going to sense that we need more time we would rather ramp up and give more resources than overextend unless there is a huge shift but everybody will >> let me add a few bits here. This is indeed 2 years but it's 2 years for the most biggest projects and definitely Paris V2 project is a big on its own and it will be started as soon as Paris V1 will be closed uh will be closed to the mainet and that part requires lots of time because it requires both engineering research and some parts that are not yet known because this is a research part where nobody in the world has all the answers. answers how it will be and it means and this makes it a 2 years involvement also to the previous questions to reduce the complexity. All those work packages that we have 17 of them are built in a way that you can understand how they depend on each other and evaluate each each one of them independently and it reduces the complexity. We tried to remove additional dependencies as much as we could and if you want to make it looks like a smaller project. So then most of those packages can be combined into one.

[00:19:53]But this way we will just not provide enough for understanding because each work package has its own matrix of success has its own understanding of we are if we are good or not and if you merge them into something something simpler then we will not be able to tell if we are successful or not and if we have reached target or not and that's our plan how we tackle all this complexity and that's why we have uh vision on the next two years and if we say just let's make uh the proposal for this single year it will mean that all the proposals will take three years not two and all of that done is to improve this throughput and improve the latency of our deliver um really good that you guys are working on all this we we do need more engineers and people that are dedicated to the ecosystem working on core infrastructure rather and simply just the IOG teams. So more teams externally from IoG working on all of this and having the knowledge and connections I think is really important.

[00:21:01]So I really do hope the community supports you guys in everything that you're doing whether it's they're building DAPs getting smart contract audits from you guys or for all this stuff that you're doing in core infrastructure itself.

[00:21:16]I know people asked in regards to progress where things are at right now.

[00:21:21]Uh c can I just get a an idea an understanding and just a recap of so last year 2025 was a lot of research and now you start the development and how far have you guys gotten into that development at the moment for example with Laos there's like a development tracker we can visually see where things are at and and a lot of people are excited about Laos having a test net implementation maybe in June this year is there anything that we can possibly look at and track the development of Paris in regards to uh what you guys are doing.

[00:21:56]There are several resources where you can check what's going on and uh we have monthly uh demos where we show what had happened and it's uh open so everyone can join after it happens we upload everything that we have on the our website that is related to Cardano projects and basically you can uh and there is also a weekly updates that we provide on everything that had happened with the project and we do it not just for Paris but for all of the projects that we involved and that we are working on. Other than that there is an open tracker where it's possible to see all the work breakdown with the dependency charts and it's regularly updated so it's possible for everyone to check what's going on. We have a discord channel where everyone can join and ask questions and we'll get replies. Also about the test net, we uh just this month we started a work on the test net and we will make a public test net with Paris as soon as it will be possible.

[00:23:05]All the Paris code is targeted main. So basically it's not something that of the main of the master branch of the projects and all the benefits that and all the features that less that we did oh sorry that we did for Paris are already available and it's not something that we will need to adopt in the future. It's something that is already there. So once the p once test net will be available uh we will officially announce and uh tell about this so anyone can try as soon as it possible.

[00:23:40]>> Okay brilliant. I did have a look at some of the resources uh that you guys gave me the uh GitHub website and all the statistics there and I'm looking at the Paris dashboard as well and playing around with the global parameters around I guess configuring it and seeing what it looks like. So there is a lot of stuff there to look at and to understand. So I'll try and put um a lot of those links and resources in the video description down below for everyone to go through as well. So there's a lot of material here and I thank you guys for constantly updating the community with um all your PowerPoint presentations and everything that you've uh done as well. We just need to get it out there for people. I think a lot of people miss this and uh don't see it. Yeah, I think that uh we have been you know we are developer company and our main thing is consultancy for global players big companies and that's what we are actually doing and at some point we said okay we are doing so much wonderful work for Cardano and let's see where this technology can be used in utility for retail for enterprises and last year we decided we're going to open a whole vertical and let's push blockchain into our offerings. Let's see which of the clients can adopt it. It's it's very hard to add blockchain to existing companies because they're scared of blockchain. It's like, oh, it's crypto and the cycle takes time. But with upcoming midnight, with privacy, with with Cardano being safe, suddenly there is a big interest, stable coins, remittance, tracking of sensitive data, hiding patient data, and our clients are interested in those topics. So what we're doing on opposite side and where we don't need the treasury is we are doing the BD work and essentially selling Cardano as one of the solution to the multiple layers of what companies are digesting.

[00:25:30]>> Yeah, we're segueing a little bit from the main topic, but I am interested in the the business development side and uh I might as well probe you guys while I have you here from from a really big development company point of view. the business development. Firstly, I want to know how long does it take to on board a client with the idea of implementing blockchain because um a lot of people at the moment think that we should build all the the scaling and all the tech or the core stuff first and not worry about any of the business development or marketing. So, it's like a chicken before the egg, egg before the chicken kind of thing. Should we start the business development now even though we don't have the scaling side of things?

[00:26:12]In your opinion, when should we start the business development? Obviously, you guys are starting now. You're talking to clients, but how long and and how long also does it take to on board clients and convince them that that blockchain is the right uh technology for them and fit for their business? So, I was really happy to be in Japan uh last week and meet the community, but I also attended some of the business events. And when you compare the need of companies in Asia versus Middle East versus US, Africa, Europe, there are different needs. For instance, things which you want to solve in um let's say in places between India and Malaysia and let's say Philippines, those problems were solved in Europe or US. Latam has its own problem. So blockchain is not a blanket.

[00:27:03]you're just going to use blockchain to solve it. It's a it's a layer in multiple tiered solution. So, Midnight is highly interesting for Europe and US because privacy for financial and healthcare companies brings them they can obiscate data and still have fast settlement fast transactions. In remittance world, blockchain is solving all the things that banking has been withholding, speed and cost.

[00:27:35]So each region has its own things and you need to start now. And it's complicated because you're not just competing against other vendors, you're competing against other chains. You have to understand why why Ripple has its own way to do things, why they're faster in some things. They're not decentralized.

[00:27:54]they have special foundation. Why are smaller some chains that don't even have footprint where they're faster in getting adoption? In the end, they need it needs to be a much comprehensive solution. You can just sell blockchain and they need to understand especially today that they're going to make money and it's not just solving a problem.

[00:28:17]It's really they're making money then they will unlock uh the vault because everybody is very sensitive where they and how they spend money today. You think about cutting cost for 10% and suddenly the price for electricity goes 20 30% or your provider of AI switches from monthly to tokens and you burn your AI budget in in one week instead of one month. There are a lot of lot of things to compensate. I would say the blockchain time hasn't passed. It's it's there but it's no longer are you using a blockchain. No, it needs to be a comprehensive solution that's integrated through the whole vertical.

[00:28:55]>> Understood.

[00:28:56]>> Yeah.

[00:28:57]>> Understood. So business development now and power through for different verticals. Okay. All right. We've we've spoken about a lot of things. We've gone through Paris. We've gone through all these integration moving parts. We've gone through the the the timelines and everything. And we also mentioned uh how to find out more about the the project and updates from you guys. Is there anything else that we need to talk about that we've missed before uh we sign off on this interview?

[00:29:22]>> Well, if if you allow me, I I have a just a closing sentence. And when we wrote the proposal, we were already thinking what will the DREPs ask. Let's let's go ahead and let's create an FAQ.

[00:29:33]And I think we passed a lot of those things already. We are in sync. We went mindful. We went mindful with time and the price. Here it's this is a one coherent delivery pipeline with Paris at its heart with scaffolding around it for safety and the tooling that makes it usable for the broader ecosystem. I spoke to Alex before this call and asked him we have users. We're not building something and hoping like people will come to the restaurant. No, there is already people wanting to eat hungry and we are opening a restaurant and we are expanding it. We're careful to separate what we can promise and what depends on somebody else. We don't want to oversell. We're doing what what's in our control. And trust us, when we're in when I say that it's it's bolded. We are not allowing that somebody messes with our timeline. We're in control because we the stakeholder is the community.

[00:30:26]We're taking this very seriously. We have been here for 8 years. the Paras and Laos are going to define the next phase of Cardano and we want to be sure once that it actually lands safely and we will be here in the in the space and we'll be happy to answer all the questions in upcoming uh days about any work package breakdown of the budget and how to coordinate with the teams.

[00:30:51]Okay. And if anyone watching this uh interview if you have any questions please leave them down below. I'm sure other DR repres will definitely have questions as well and I'll pass them on to the team here so they can answer them clearly and uh as promptly as possible.

[00:31:05]But Chris, Alec, thank you so much for joining me on this episode. I learned a lot about what you guys are up to and some of the tech here as well. It's always interesting to get some of the engineers on the podcast so we can learn about this stuff. But uh anyone else that's watching, if you learned something as well, please hit that thumbs up, like, subscribe, notification bell on your way out. Make sure you follow the Twe socials and all of their links and everything, all the references down below so you can learn a little bit more. And check out that proposal as well. It's probably one of the core infrastructure ones that we do need to vote for. Do check that out. Again, links down below. Alex, Chris, thank you so much for joining me and we'll talk again soon.