Fireside Chat With Peter Kieltyka and Kartik Talwar

For the next talk, I'd like to bring on Peter. Please give him a big round of applause from Trails. [cheering]

Hello.

Hey, everyone.

Welcome, Peter. Uh, we've known each other for almost a decade now. And, uh, you've also been building in this space since then. That's how we got to meet each other. We uh we had the same office in Toronto.

Um Mike, the first question to you is understanding what got you into this space at the beginning and um do you think those same values are held still uh by you and your team?

Yeah, it's been quite a journey. Um likewise seeing your your amazing journey as well. Um you know what really caught me into crypto um was really the first time seeing internet native protocols around digital assets in internet native finance. Um for me it was just an massive design space that was so impactful that I wanted to get involved in. I mean just a combination of distributed systems, cryptography and economics really represented in me a new organic structure that I wanted to build around.

Um what for me what I always wanted to do was recognizing these protocols were were essentially discovered and invented and evolved um and now becoming improved. I wanted to really drive the utility around them, right? I wanted to build applications so like my friends could just use this stuff, you know, [snorts] like how can I make this stuff so easy while maintaining its integrity. So just my friends could benefit from from this from this thing. And for me as a person who grew up on the internet, who loves the internet and I love open source.

I've been doing it for for multiple decades. Um being able to this is my chance. This is my chance to do something cool on the internet because it's finally something new here. This is the time. So it I mean it it there's still incredible opportunity and there's incredible impact ahead.

So I'm just as passionate as ever. Yeah,

that's awesome. Uh, it's always good to hear that the energy is still there. Um, maybe just to kind of set the stage, tell us about how everything started and what led to actually starting building trails.

Yeah, it's uh it's been a fun journey where we back we started in December 2017, but of course we kind of I knew I my next startup was for sure going to be in crypto or my new my new invention. I love just making stuff. Um, we in terms of we wanted to build a product. We looked at the low level. My background is a lot of infrastructure back in engineering um really and full stack, but that's where I where I really specialize.

And so I wanted to go from the the raw metal right to the user to deliver something they could do and really cool. So I was like, I want to make a video game. And we actually built uh we were set out to build a trading card game, Skywever, some of you I'm sure know it. And um there's such a massive design. So by January of 2018, we realized, we can't use MetaMask.

Like we have to build which is you know it's a it's a really robust wallet that's stood the test of time to this day but we have to be able to deliver an experience that any user can use any average user which means so many different challenges and friction points as simple as like signing with a with a native social credential or having dealing with gas and then where we spend so we knew by January literally a month into we're going to make a video game or like now we're making a wallet all of a sudden you literally um and then from the wallet it's like well how like and also looking well how are chains going to be scaling what is the evolution of these chains cuz clearly you can't go do this on layer 1. You know, there was such a great amount of global research happening around scalability, which is how we got to meet your crew, right, and all the people around them. And we were just very fascinated um by that work and seeing how things were going to happen because it informed, you know, how we were going to impact the application layer, which is really where our our we're we're centered to focus on. Um so again a lot of the work that we do and we've been focusing on is really to accelerate the utility of crypto and web 3 and for us that is at the application layer through you know and sky was really cool and then we really it was always given my background sequence is always the baby for me you know behind the scenes baby and it's a very large platform um where we also spent a lot of time working on smart wallets you know as early as 2018 and uh yeah so we just

before account abstraction all those things were even like a fully formed concept.

Yeah. Nobody was Yeah. Like um because it's what was necess like we needed to solve the gas problem. So we needed metatransactions but you know what's better than metatransactions is a wallet that can do it natively which you know which is a smart wallet and can also enhance security and usability and all these. This was like yeah so we were you know actually our team even published in 2018 of September uh the ERC 1271 right so um you know we were very early on finding the right primitives and the building blocks that layer on top of these blockchains to really again accelerate utility and that's what you know continues to drive us.

So we're how fast forward to trails um you know it's been about the application layer it's been about the developer and it's been about not just the one product we can make but all the products that everybody can make and really servicing all the different developers and and unlocking um the challenges that are in front of them so that they don't have to keep you know solving all these little micro challenges and there are there are so many um and even just quering data on chain like what are all my balances what do what do I what do I have or and then and then do that times 100 chains you know um or being able to send a transaction when you don't have a gas so where it led us to It was very interesting. We we sequence is a very robust platform that's evolved over all those years and you know and it continues to stand to stand strong and it's really the same architecture and primitives evolved. It's it's actually in its version three at this point. Um and we've continued to to maintain its robustness. But what the biggest problem for these app developers is actually making sustainable viable business where they're making money.

You know, so it's one thing like they work their hearts out to make the product to you know get the grant get the incentive make the thing but now you have to go make they're actually building a business. they're just not making a fun app, right? Um, and so to do that, they have to monetize. They have to be able to actually make money. And when you can create that opportunity to make money through sustainability, then you get this effect of like the Apple like App Store gold rush.

You know, you can make a fart app and you get rich and you're like, I'm going to go make a bunch of other apps too, you know, and everyone and suddenly that is driving utility and but so the the fundamental issue was just the the natural evolution of blockchains and on their own which was there are many chains now. there's, you know, hundreds and and that's fantastic and but they also have even thousands of tokens and we have dozens of wallets. So you have this insane fragmentation problem, you know, of like you actually have to have all these things combined in order to actually have a user show up with on the right chain with the right token and have a gas token so you can actually take their money or receive and transact with them. And that could be any kind of an application. It could be, you know, it could be a DeFi thing, it could be a it could be exchange, it could be a prediction market and chains similarly need the liquidity and the value to flow freely through their network.

So the fragmentation problem what I say the triple fragmentation problem across chains tokens and wallets and even even guest is so significant that it literally handicaps anything that you can do. So this is to me literally the most important problem and that's what trails does um effectively just by being an interoperability of across interchain and inter token you know orchestration layer that allows you know users to have a very seamless experience of doing that and interestingly trails um although a separate brand separate product 80% of the technology is literally just sequence under the hood um and I can talk more about that. Yeah.

No, let's let's jump in. So, um, Trails is obviously a more developerentric product and, um, well, first, also, congratulations for publicizing everything today. You just launched today, which is a big moment. Um, it also the right timing because we're seeing so many new chains form. We're seeing liquidity kind of just getting fragmented.

And there's a lot of things that are exciting as engineering problems, but also there's a UX piece of it and there's also the third side of it which is like how do you actually make sure that the thing gets done and that's kind of the focus not how or where or when. Um so the trails website says this is a oneclick sort of solution to uh to kind of bring crypto rails and simplify that. Um break that down for us. So what exactly is happening when somebody use your SDK and what do you simplify kind of for for them?

Yeah, I think yeah for us um where our teams you know we have a lot of very passionate members on the team who've been in the space also leading longer than me right now. Um and not everybody but some and um you know for us it was always a vision of Ethereum as one. You know Ethereum is not one chain. Ethereum is an ecosystem of chains and how do we have the system be able to work as though it's one as though it's one token to have such a seamless experience. Um and that's where the one click crypto rails kind of the notion comes from.

Um in terms of an SDK perspective there's a there's a number of layers right there's of course our intense primitive which is an onchain way of representing a set of transactions that can be you know merkelized encoded into an intent address and then used to encode and execute across multiple chains. Um, I'll get into some of the details on how that works, but what's neat about that thing is actually an account abstraction based address and users just deposit via their one-click, you know, confirmation from any machine and then it'll just kind of fully execute. So, I'll get into more of the the specific, but the SDK is there's a number of SDKs you can use the you can the contracts are open source, the client side SDKs are open source. you can technically use this onchain orchestration primitive intent primitive on your own to really execute any arbitrary kind of transaction that goes through it. Um which of course is still a lot of work.

So we package a client side SDK with trails to give you a full widget that you can quite simply npm install 0x trails um connected to any kind of connector. So if you have you know if you're using connect kit, rainbow kit, wgme anything at all or we'll give you a connector. You can use any wallet as well. So you don't have to change your application to you don't have to use a smart wallet. Although as as much as we are an advocate of smart wallets, you don't have to use a smart wallet, you can use MetaMask.

In fact, you can use MetaMask without even 7702, which is so cool. And it's one of the no like one of the novel properties of the print 10 primitive that we've created. And so you just really add it to your your wallet. Uh sorry, you add it to your application and then it will just you know essentially prompt the user what kind of what's the origin token. So you you the application will say in all cases and you know an application will construct a transaction that it wants to do such as to this contract address maybe send a value maybe call some function on a particular chain and so that is effectively like the app's intent right it then prompts the wallet and says hey wallet can you ask the user to sign this and confirm this thing and then it'll submit it to the network that is the real the intent of what the user is looking to do and the application wants them to execute but But in many cases due to fragmentation it fails.

It's in you know insufficient balances and so forth and so because of all the different challenges. So what trails does is it really works it treats that as the destination intent and it encodes backwards in reverse to prompt like the destination address and then it goes from the origin. So it prompts the user with the sequence indexer showing all the different tokens that they may have and the user can just pick any origin token they want seeing the balances and the values and they can pick and then once it does that it then talks to the trails API which is actually um really job is to solve and find the optimal route to go and get the token to the right currency across the bridge and then execute on the side and combined with the intent primitive which is again a Merkel set of operations that's encoded it will then just there's a single user confirmation user deposits into the origin address. Sorry, there's so many details here. Um, docs docs.

trails.build. All of this is documented as well. Um, and then it just will fully orchestrate and execute. And what's neat about the the system that we've created um is that it's fully on chain.

So everything is transparent and any you can forward it on the client side. So you can once you've encoded and you've transferred funds into the deposit address, it is really the full set of transactions that are um you're doing. So it's very anyways it's a lot lot there to

I can see Peter is excited cuz uh he talks faster when when [laughter] there's a lot of things to say. So So let me let me see if I understood it. So, one of the the benefits that we usually get pitched for different products is like, hey, if you don't have the gas token, um you can like pay other tokens or we'll like pay it on your behalf. That's kind of like table stakes and kind of the most common example we've seen. But what you're saying with Trails, you can effectively say I need to pay this merchant on the other side 100 USDC, but I have no E or USDC, but I have all these other tokens in my wallet and also on different chains.

like you kind of combine that and you'll figure out how to get all of that to be 100 USD CSTM and that's what the other side gets.

Yeah, very very well said. Thank you.

Amazing. But all that is through intense and through multiple chains.

Yes sir.

That's awesome. I mean um that simplifies a lot of things especially accounting um because we're in this problem all the time. Uh we kind of let people buy any of our event tickets in crypto. we have to think about how do we simplify a lot of the receiving end and think about tokens and just kind of managing all the different chains. Um the other question I have for you is what else does that involve?

So a lot of the intents end up being super flexible. Is it just applicable to transactions from sending tokens A to B or do you have the ability to kind of do arbitrary logic?

Yeah. Although like a significant problem is crosschain, right? Because you do need to swap, bridge, and then execute. And it could even be a transfer. Maybe you want to do a crosschain swap.

Maybe you want to interact with a contract in a different place. Maybe you want to mint an NFT. Maybe you want to buy an NF thing like all these any arbitrary literally any arbitrary transaction. Um, but what's also really handy is just the same chain. You still have fragmentation on a same chain basis because what if you don't have the right token that you the contract is looking to to interact with?

That's the other problem. So we and the other problem is well, what if you don't even have the underlying gas? What if you like stables are a really big topic. What if all you have is a stable like USDC or UST or something else and I don't have any other you know option to to pay my gas fee. I don't have ETH or pole or whatever I may need on a on a on a native network.

So you can actually with the system with Trails we kind of it kind of comes from like this notion of transaction rails right we kind of enhance we enrich the transaction model.

I just understood the name.

Yeah.

That's where it comes from.

Transaction rails trails.

That's what we did. And you know and it really you could use it for anything. So we have you can use any ERC 26 uh 12 uh permit based ERC20 token. So if they have a permit option operation you can go and just assign you know assign a 712 you know payload and you can literally have it instantiate do the deposit which will then do the full orchestration. Um so aka you can have a full stable coin end toend flow.

All you can have is stable coins and you can do everything you want and if a merchant says I want to be paid in ARB or uni or pole or or anything or some other fartcoin you can get paid in fartcoin you know no problem as long as there's a dex and there's a liquidity route you can do that

so the API figures out what is the best path to that token and manages the the selling and the the arbitrage or rather just convergence

that's right so essentially yeah there's the like I mentioned the intent primitive which is really encoding these transactions onchain but then the job of the back end or really any back end, anyone can make a different backend um is to find the encodings, right? The the set of operations, the set of transactions that are then compacted into the intent address and that's how it all works.

From kind of just the the UX side of things, how fast does this experience?

Yeah, that's a great question. It's quite good. Um I would say we average around 12 seconds, but I've seen it 8 seconds and I know we're going to get it down to five. Um

and this is like assuming like only one chain hop or doesn't matter whatever the goal is it ends up

so that's a crosschain full yeah that's a swap crosschain execute

I mean that's that's pretty reasonable size of a one confirmation on on mainet um and it just kind of managed everything you said about

conversion paying for gas bridging and any of those scenarios are accounted for

yeah in in kind of the the the longest case you might have a token on the origin that you want to swap and then you'll bridge and then once it reaches the destination you're going to want to execute you know and whatever it may be. So um and the neat thing about our design is that the although like I said it works with MetaMask but what's neat is the intent address is almost like a ondemand smart wallet that is owned entirely and controlled by that wallet that deposits in the first place and encodes in the first place and as a result via this account abstraction stuff we essentially en enrich MetaMask to or any EOA wallet or you know any to to act with via account abstraction. So suddenly have bundled transactions. So now you can actually bundle the swap and then the the bridge as one onchain transaction and then it's another transaction on the you can technically do a third hop too um if you need that many routes. Um, and so in some ways, if you think about it, it's the same thing a user would do manually if you really really had to.

Like that pain, right? Okay, I want to do this thing. Oh I don't got the token. I Okay, and then I got to go to the the unis swap. I got to swap it and then okay, I got to find the right bridge and like bridge and I wait and then I go and then I do the transaction.

That's literally what it's doing for you, but it's just doing a way better job on finding the optimal route. And you're doing it in a single click and then it's just executing it for you, finding the optimal price, lowest latency, and a few extra tricks with the gas stuff.

That's awesome. um kind of zooming out and just kind of looking at this category a lot. Do you still feel like there's some general infrastructure pieces that are missing on um how we can actually simplify crosschain? Like this is kind of your wish list of what the chains would do better or what other systems can uh can do to simplify building a product like Trails. Yeah, I would say um look the application layer of blockchains and Ethereum is is very powerful, you know, um more powerful than people may realize, you know, like we could have built trails three years ago actually.

Um we didn't need I mean there wasn't as many but like it's becoming a very big problem now, but it's it was possible 3 years ago. Um I think my favorite thing of all, my big Santa wish list would be having this native account abstraction. Um I think effectively trails is by leveraging the power of account abstraction we're able to do this this awesome experience even the Ethereum interop layer stuff that was you know announced a few days ago which is really cool that like a 4337 extension for doing interop 437 wallets and you know doing you know s working around similar goals to trails but not supporting EOAs and metam masks and all these things so there's a bit of a different thing but as you can see they can achieve what they do via smart wallet so I think there's nothing more important than a protocol I think there's all these different like from a realistic path to engineering to execution to delivery. It would have all and and one of the great things we have about all these E the Ethereum ecosystem is the compatibility across these chains which is just so fundamental

and is just commenting on just they all use EDM is that what you mean or is there something more what does compatibility mean here? Well, just you know the reality like if you're like back if you remember the evolution of L2s like not everything was by code EVM compatible. So that that creates even like even your programs don't run the same then you're totally toasted and then developer tools don't work and it's like even developer fragmentation is yet you know so having we have today much you know very strong like sequence on 50 plus chains chains understand how important compatibility is to maintain even compatibility at by code level even zkync has been doing that like you know they've made some great strides towards that which is you know super impressive team um but like again back to native account abstraction we all want it we have 437 which is really cool but like you know it's going to go through version 0.910 breaking which is cool but you know still centralized bundlers blah blah blah blah like we all want native account abstraction we don't want some random chain making native account abstraction and when we have it we literally

some L2s have this natively but doesn't then you run into a bunch of

nobody wants I don't want that you know what we want is everyone to use the same native account abstraction it will literally elevate significantly to the next level of the application layer theorem and all the things that you can do and I'm sure the developers

how do we get there is it just like we just effectively complain until it's there or do you think this is just an engineering thing that uh requires problem solving thing was purely we just have to make sure people agree on it and and they're not right now for some reason. Yeah, we just got to ask Vitalik and you and that's why I'm here to tell you to tell you tell Vitalik, you know. Um, but I'm going to call him right now.

Call him up. Um, but you know, he's been a huge advocate of kind of he's been singing for years, right? And so, luckily, thanks to him, we've been making great progress. I mean, 43 through7 of course is is a significant um direct, you know, directionally where we're going. It's, you know, there's the EIP7701, which effectively looks like to be the native abstraction standard.

I'm sure it'll evolve and there's learnings and I know it takes time. So, like everyone's doing the work. I'm just saying please make it a priority. You know, I would love if the sooner we get it, I think the sooner we everyone wins essentially.

No, that's a that's a that's a good way to put it. And I I'll text him after.

Um and I guess like lastly, uh what do you want people to going to do with Trails? How can they learn more? And uh what's kind of something that you're looking forward to on uh people building on platform?

Yeah. So, you know, please check out our website trails.build docs trails.build. We have a telegram.

You can come chat with us. You can see your contracts. We'll give you more details on how it works. It's easy to integrate. It's free to integrate by the way.

Just we don't there's no SAS fees, monthly fees, none of this. It's all batteries included. Like everything is like all the sequence infrastructure is behind it. Literally no access, indexer, relayer, all the thing yada yada yada. Uh very easy to integrate.

Although I may maybe I made it sound complicated. I just want you guys know how it works. Um but um very smooth and um you know really and I think everybody faces this this this problem. if you're on some chain and you want to just get paid and or have people interact with your thing like this, you know, I think just give it a try. And we're also here to work with all of like everyone.

So like we want we're we're developer tools platform and company. So um it's a priority for us. The road map is the road map of our customers and our and our you know our partners. So um yeah, we just want people to build cool and um we're here to help.

Trails.built. Thanks Peter. Please give him a big round of applause.