# EIP 7702 and the Programmable Wallet Era panel - Bogan Habic, Franceso Andreoli, Mirko Garozzo

- Speakers: Mirko Garozzo
- Channel: [ETH Belgrade Community](https://streameth.org/eth-belgrade-community)
- Date: 2025-10-07
- Duration: 46:21
- Topics: People & Blogs
- Watch: https://streameth.org/watch/yt-FlRqGtq4zTQ
- YouTube: https://www.youtube.com/watch?v=FlRqGtq4zTQ

## Description

EIP 7702 and the Programmable Wallet Era panel - Bogan Habic, Franceso Andreoli, Mirko Garozzo | Tenderly, Consensys, Zircuit

## Transcript

You guys hear me? Test test test. Awesome. Okay, sounds good. Let's welcome our uh our panelists on stage so we can actually like let them intro themselves. Mirro from Cirquit and Bogdan from Tenderly. Thanks so much. Round of applause here. >> One, two. >> So, how many developers we have in the room? Raise your hand. >> One. Okay, sounds good. Awesome. Awesome. Awesome. Okay, let's do a quick round of intro. Bogdan, you want to start? >> Sure. Hi everyone. Bogdan, um, I'm the tokenwide guy of this conference for a bunch of these panels, I think. Uh, I'm one of the co-founders and CTO at Tenderly. Uh, Tenderly, for those that don't know, is a fullstack infrastructure platform spanning like four different product suites. Uh, I'm not here to shill Tenderly. I'm here to talk technical details. So, at some points I might talk about how Tenderly looks at these things, but that's me. That's enough shilling of tenderly itself. Hey, I'm Mero. I'm I work for Zirk where educator lab. We heavily focus on defy and we also done actually a lot of work around and accountraction and so on. Yeah, >> nice. What's up everyone? I'm Franchesco. I'm leading developer relation at consensus uh the umbrella company around MetaMask in Furalinia and yeah super excited. So this is all about kind of postpectra. Let's kind of like give a context to everyone. So um what happens around kind of like developer experience prepra and postpectra? >> Okay. So maybe uh a question for the audience so we get some engagement. Who here logged into a web3 app via Gmail or GitHub or whatever? So a few of you. Okay. So that was done. You probably went through privy because they feel like they're the market leader for this. But that was uh 4337. So that was a way to deploy a smart contract on chain that behaves as if it's your account. There was a few problems with this. One of the problems is is that every time you log in, you basically create a new account. So you're fragmenting your own liquidity and it's very hard to keep track of the keys. There is ways to tackle this, but basically that was what happened. Now in the postpro world to answer your question concretely, what is changing is we can talk about a bit later about the off call and what didn't happen. But here uh Ethereum added the new transaction type which basically gives you superpowers or another way that I like to say it is it stole the idea from meta in Libra actually. So the idea is that you can turn your uh EOA, so your normal uh normal um account into a smart contract for the sake of that one particular transaction and that's it. And this is basically gave developers a way to run code as if they were you without actually having you deploy a smart contract that is uh impersonating your wallet. >> Yeah, I guess I give you a bit more kind of background contest. So there was another EIP 3074 that it was around for a while and there was a lot of drama I guess in the community should we use 3074 should we just go fully for 337 and that's why one day Vitalik out of nowhere I guess in 20 minutes he probably brought the 7 77 O2 spec and then we decided that that okay 772 is the way to go and the cool thing about it is that it's kind of a hybrid because you can still full integrated into 4337 So you can have like kind of the best of both world when using your wallets. >> Yeah. Think think about like the the developer experience today like what are kind of the main pain points that you think developer has while building their first DAP. Uh for example yesterday we announced the web3 off acquisition. So embedded wallet is is obviously a big uh big uh kind of like direction. Um I think like on boarding users is one big things but also like managing the developer expectation is also one important things. We're going obviously in the direction on the CLI side. What do you think are kind of like the pain points today around the around developers uh developer on boarding but also like developer tooling? >> I think it's completely different how people think about uh authorization and authentication in this space compared to web 2. I think that's the main limitation because the usual thing that web two developers go through is something called oath. So this is the login with Gmail or whatever and there you get a particular token that you use for the rest of the session and you don't necessarily think about these things. So you only do it once. The screwed up part about web 3 is that anytime you want to actually do anything uh I like to say that you go again through the whole flow because you need to ask for access, you need to ask for authorization and all of this stuff. So I think part of it is like is the mental burden of like changing how you think about these things. So I think that's the piece for for onboarding developers. And then also in general when you're a DAP developer, it's overly weird also and I think you see this more than than we do like what's the job of the wallet versus what's the job of the DAP because it's weird. Everybody expects the wallet to protect them and give them capabilities, but it's also the DAPs wanting to wanting to have those capabilities. So I think right now what what is inevitably happening is at first we're going to have curation similar to what you guys are doing with 7710 if I'm correct. uh which we can talk about later as well. By the way, we're going to use a lot of numbers that mean [ __ ] nothing out of this outside of this room, but I think it's going to be fun to talk about. Um and then I think we're going to start at first with a lot of curation and actually in a very constrained system to what it can actually be and then in the next six to 12 months we're actually going to see what developers build with this and which developer experience actually pops out of this. right now. >> I guess uh also kind of the tech how to grow the UX u since you work at MetaMask. When MetaMask was started, obviously it was a wallet for a single chain back in the 2016. It was about uh only focus on mainet and so were most dups at the beginning. But then we started to have like one gazillion L2s. So we started to integrate it on EBM chain and for from the UX side from for both wal especially wallet but also D developers it's has been like very complex to kind of to keep up to date because now we have to think about bridging and we to think about paying gas in different chain we have to think about the gas m we don't want to pay gas in a native native token but we want to use a different token so I guess that's was one of the main pain point it was all the complexity that has come since 2020 until today. Yeah, I think like stuff like also like chain abstraction uh it's also like playing a big role but but I mean going back a little bit on the 772 stuff like so what actually actually happened because somehow uh I was was happy that we kind of like ship this part post spectra but somehow also like I tried to stay realistic about I started to see obviously the b like bad transaction back revokes that I thought okay that's actually super cool that you a developer can revoke 100 accesses to the apps in in one time but and then I thought about a more realistically preview I was like oh but in web two you know when you start doing batch transact and you kind of like signing multiple transaction it kind of like was a minimal expectation for the users right I think we are we are getting there uh but what else can you actually do around this kind of like post uh 7702 uh with um uh on the developer side also on the user side right I think you can enable a lot of different new features but maybe you can deep dive a little bit on on on what's happening after >> sure so Uh all of this technically isn't a new idea. I think why this is innovative is because it's much easier to do. I mean that's the that's the main thing. Uh you had similar stuff that you could do with intent by the way. Very similar to signing an intent and then it would do something. It was a bad user experience but you could you could still do it. The case that you mentioned which every article mentions here because I guess that's what chat GPT coughs up for the person writing the blog is the approve and swap. So right now if you go to unis swap for example you know that you will need to approve your tokens preact you would need to approve your tokens that's one transaction and then you would uh you would swap the tokens and that can either be an intent so something called permit too or it can be an actual swap uh happening happening on chain. So that was at least two transactions that that need to happen with EAP7702. What this basically allows to is for the sake of this particular transaction or this particular interaction maybe is a is a better way to put it. Uh what unis swap can do is turn your ea so turn your wallet into a smart contract only for that particular transaction and basically do the approve and the swap in solidity or whichever other language they want but in solidity do the approval and a swap and as a user this basically means uh this basically means it was a single transaction. more concrete example. Has anybody ever signed here anything over do you sign? Okay, so this is very similar to imagine that you have for example I'm the signator for tenderly which uh I I I really hate but that's the fact of life. So I have like a couple of things that I need to sign every day sometimes multi a lot of them every every day or every week. This would basically mean that if I didn't want to read the contracts and we're going to talk about security I think the doign analogy will actually go a long mile during this panel. Imagine that they can go select all of the uh all of the service agreements and contracts that they need to sign click one sign and all of those actual contracts like real physical contracts are executed or signed by me. So this is what 7702 does this analogy make sense if it's bad sorry I I'm a bit sleep deprived I really try hard to to make this one work that's one piece but then also what's possible in web 3 uh paying in different tokens so not paying gas in ETH but actually paying in something else so uh it allows for that uh I think that we will see stuff like scripting uh which is a case that I didn't see mentioned uh this was actually something that we did as a hackathon project at E Berlin it was called sold. TTY but uh what we can have as a developer experience uh you guys know how you have like a ripple like a rede valid print loop where in python you can write something and you can execute line by line you will be able to have if somebody wants to do it a similar experience now for solidity so you could actually write a piece of solidity send it onchain and it could be executed as part of your EOA nobody talks about these things it's questionable if you would like to do that but I want just to explain like the flexibility of stuff that can actually happen for these things. Um, do you have any more user facing >> social recovery? It's one >> social recovery. Yeah, >> it's a big one especially right now with UA serbas. Great. But like there is a few kind of cav around the UA. One of them it's obviously like if you lose your uh your key, you pretty much lost access to your wallet. And if you think about know even if you compare to traditional banking if you like if you lose your credit card like you still have access to your bank account right and I guess that's what 33377702 kind of wanted to get closer to like even though you can still lose your uh your key but you still have a way to access your funds you still have way to sign transactions and so on. >> Yes. Yes. So social recovery I think is a big one. Um I actually recently started to see like some some hacking happening. Uh I don't know you touched on security. I think like we should kind of like explain a little bit what's happening in the space that people were saying okay people got hacked because of kind of like some uh delegation of uh of accounts. Um where where do you think are kind of like those uh those are there some security issues? Uh because I kind of like trying to uh follow a little bit but I also didn't understand what level of kind of like breach happened. if it was on the human level, on the delegation layer um and or even on the DAPs developer layer. I think this is just getting bigger, more we abstract also more we opening entry points. >> I think so the largest delegation or the largest u hack of this type is the bybit one. I mean it was done through the safe UI being hijacked but what they did is they changed the implementation of a safe. So a safe is a smart contract. In this case, how that could happen? Imagine that instead of a multisig, they had a normal wallet. They would just change into which smart contract this wallet turned into or like which smart contracts capabilities the wallet assumed at at the time. Um my prediction, whatever the [ __ ] does that mean, but my thoughts on what is actually going to happen is that we're initially going to see two broad categories of hacks that are going to happen. Uh, one is people just blindly signing delegations and people draining funds from them. To be completely honest, I think what MetaMask is doing, and I'm not not shilling just because we're on the same stage of not allowing you to delegate any uh to anything unless it went through like a rigorous process is a good thing because if you let people to delegate their access to their account to anything, they will lose their funds. I think that's like one broad which for a very long time will cause issues unfortunately. The second one is uh probably certain uh similar to when how uh flash loans came out like nobody thought about what can happen and I think that that whole second category of problems is going to be my DAP was written to work with EOAs and 4337 wallets and I really want to adopt this thing and then somebody's going to figure out some weird way to do re-entrancey or or something else and do those types of back. And then I mean from the second category, what we're going to have is standardization popup. Unfortunately, we're one of the space that uh spends most money on standardization because we need a $600 million hack for someone to write an ERC. And then from the first C and that was a joke, but unfortunately it happens. And then for the first category of things that's probably going to happen is going you're going to have providers like us or blockade or cyers or whoever uh do scanners that are then embedded in wallets like meta mask zerion ledgers of the world or whichever and try to protect users in in that way. So I think those are like the security implications in the short term that's going they're going >> how granular can the delegation model be. Yeah, I um so this kind of touches upon 7710 I think as well. Uh are we talking about session keys now or I'm just looking at that to say who knows what the session key is? Yeah, >> cool. Who knows what the cookie is on a web two page? I'm I'm guessing more people. Okay. So, other than the overly frustrating popup that that happens on on everyone's screen, cookies is a way is you log into a website, you get a cookie and then you can do stuff until that cookie expires. So, similar to that, we can look at session keys as well for 7702. How does this analogy go is basically uh I can delegate uh my EOA, my wallet to a certain smart account, but I can uh basically think about three things. I can say what so which functions can be called. I can say when which is basically is it in a certain context as in uh three times a day or until next Sunday or like those types of uh things and then finally it is who who can actually do this on my behalf. An example of social recovery can be only these particular people can recover my funds if I didn't have any activity in the past three months and uh they can only send funds to a to another EOA and that's it. So those would be like the the what when and who for those types of things and then basically session keys give us a capability to have granular permissions granular permissions for these cases. I hope that was straightforward. No, super interesting on the on the action keys. Um question on the on the infra side. I mean going a little bit lower, right? Like obviously like people uh at this point know what bundles are, what what is an RPC. Um how this kind of like infra is moving uh forward to kind of like this postra era even like the on demand pool you know entry points like I think there is a lot of interesting discussion here. >> I'm not flashbot so I cannot answer to the pool stuff. Uh the straightforward stuff is there's a new RPC called called get wallet capabilities if I'm correct. I have so much useless information in membrane. Sorry. So it's called get wallet capabilities and the point here is for ADAP to figure out what the RPC can actually provide in these uh in these cases. So that's one thing I think all of the AA providers, the alchemies, Pimlicos, Dynamics of the world also will need to ship stuff either different entry points, kernels or however they they call them to to support these delegation patterns. So there's a bunch of that that needs to be built out and at least to my knowledge that is mostly not actually built out yet. it's only just a a few things built out and then um I think simulation providers like us like right now we will simulate and return you results that are useful for a 7702 transaction but I also think we can do much more useful things with the with the things that are happening there both from a security scanning perspective but also then from like a developer tooling perspective too for you to reason about which delegation happened to whom how did the state change and and all of these things >> changing uh bundlers are kind of not as important. >> No, I think they will. So, this is the AA piece that they mention. I think pay masters will somewhat as well evolve I think. Yes. Although, but I'm not an AA provider so I know the surface level two blog post deep implications for >> makes sense. Yeah. I mean um around tooling, you touched on developer tooling. What about like on the security tooling side? Do you feel like these need to evolve or what are kind of like the gaps that we need to kind of like cover? Yeah, I think I mean the most straightforward scanners will be uh prevent the user for being the first one to to delegate to a particular smart contract. That can be one. So I think it's much easier now uh when you're interacting with these. I I don't envision a world where we have hundreds of thousands of these smart contracts that we delegate to like most of the usage is going to be like unis swaps a proven swap. It's going to be a thing on poly market whatever they call it like it's going to be only a few. So I think it would be fairly easy to detect and tell people are you actually sure that you want to be the first one to to delegate to these at least that are my thoughts but what are yours I think you see this as well a lot >> yeah yeah no no yeah we're seeing different things I was wondering more around uh uh you know also like tenderly moving forward because you guys support fast iteration with like loggings and and stuff >> yeah okay uh I mean that's the okay so I wasn't maybe straightforward with the answer so for us I think the most apparent piece two pieces are first the security aspect. So recently we shipped uh if you have a ledger device with a touchcreen now when you're signing a transaction we are actually one of the providers that will scan that transaction you can get like a security report for us that uh where you can see what will happen like will you get rugged will you not get rugged I think a bunch of work in that direction is what's going to happen and then the second one is inevitably something we call virtual test nets which is like a testing product and this is the shilling part sorry but it's I think honestly the best way to test daps and to test wallets connecting to a network uh uh there I think helping people reason about what these delegations do and like in which state the EOAs can actually end up in what are the contract interactions because it's weird it's an EOA that's behaving like a smart contract that might be interacting with something that's a proxy they can be like very hard to reason about so I think most of our work would be helping developers not screw up during development and then helping wallets and the last mile delivery problem of preventing users of like shooting themselves in the I think it's also I guess very important the standardization part of it because now everyone is kind of rushing to 7702. Yeah. Especially wallets. >> Yeah. I think everybody's trying to ship stuff. Uh this is why I like the Metamas thing of like opening the floodgates slowly versus yolo do whatever the the [ __ ] you want. So I I think Yeah. Sorry. You're getting my full personality. I swear >> to me. But yeah. Anyway, uh the the whole point there is yeah during this rush people are inevitably going to get rugged but what are your plans maybe at Zirket? I think that can also be an interesting one. >> Z is interesting because we have cover the security of I'm a huge fan this huge security model and we have this kind of machine learning already implemented that it can detect hacks and kind of stop malicious transaction. So what we could do is that let's say I don't know an example if you are on a unis swap LP and there is like a kind of very weird event where you kind of draining most of your liquidity we will kind of quarantine the transaction until then there is a kind of consensus mechanism around okay this transaction was malicious or not and I guess that's very cool with the delegation part as well because we >> I think and it's happening on the sequencer this is the thing that they really like about Yeah, it's a sequencer level security we call it. So I I guess it's very different from what other two do which like most L2 kind of focus on scalability and it's like for us it was like we are a very technical team and we are very security focused team so it's that's why >> did come out of quantum so it also make makes sense. We don't say that loud. Yeah. >> Okay. What? Okay. But yeah, but technically but technically I think what is super cool is okay. If we have wallets as the last mile delivery, the user still clicks yes sometimes. >> Of course. Yeah. >> Maybe too many times. The user still click yes. Even if you tell them you shouldn't do this. Really, please stop. You're losing your whole life savings. You will want to off yourself if you click yes. They still click yes. Uh I think this idea of quarantine of like the sequencer protecting you as well as like a guardian angel is a very interesting very interesting uh approach to this and I think it's also a very good way as like everyone's rushing you guys can on a network level enforce the slowness as a feature. Yeah, I guess the downside is kind of the censorship level, right? That you could argue that it's like it's a bit of censorship added, but also it's like by using an L2. I guess there is always a kind of centralization part no matter what L2 you're using, right? >> I think it's that I also think I have the unpopular opinion that Sony didn't screw up anything with censoring on Sonium. Uh what do you mean by this is it's their own chain. They can do whatever they want. Uh the true decentralization is that you can choose not to transact on Sony. That's the decentralization piece. If someone wants the so if someone is okay for Zirket to incorrectly protect them, which could be interpreted as censorship, they can go transact somewhere else. If I'm uh managing hundreds of millions of dollars and they want that safety, but they also want access to DeFi markets, then I'm going to transact in circuit. I think it's very straightforward. Decentralization is a choice. It's the choice itself, not the centralization on each network. >> I mean, it's the same with with MetaMask and the RPC viewer, right? You can choose to to point your RPC >> to another one. Yeah. Yeah, that's >> it that makes sense. >> What is Lia doing around >> is doing different things. You will see a bunch of announcement coming in the next couple of weeks. >> Yeah. Are you in Berlin? You're coming to Berlin or ECC? >> Uh ECC. Yes. Yeah. >> Nice. Yeah, that will be fun. Yeah. And um yeah, there are different kind of like levels. Um yeah actually like couple of announcement coming on the on the prover side uh but uh that would be that would be interesting to see but in Berlin we will speak more about that. Um so let's let's move a little bit more higher level. So actually what is happening now? We shipped kind of like the postpectra era. What is kind of like the could be research around EF also like what you guys are thinking? What is happening next? Like is there any EIP that we should kind of like uh focus about or like more high level right? What are the next kind of like big milestones that we should ship? I'm just hearing like Vitali a couple of days ago mentioning like what was that 10x in in the next year in terms of like u kind of like yeah aside yeah approving all the transactions I was like okay that's interesting um but what practically speaking what are the next kind of like milestones that you think like they need to be accomplished >> so the interesting thing is for mera felt and the the the team that's doing VM stuff in our in our in our company is going to kill me after this but It feels like it crept up. What do they mean by crept up is that the final stuff that's going to end up in Pekra kind of made it in last minute. I don't know if you guys had a similar feeling, but it is the largest upgrade ever that happened to the Ethereum ecosystem because you had uh the blob stuff that was shipped. You had uh this so type four transactions that were shipped. You also had uh on the consensus the staking uh the staking thing where now people can um because before you would need to stake exactly 32 ETH per staking node. Now you can actually stake multiples of that through a single node which is a very weird balance to get like for centralization and some other stuff. But it's it was the largest upgrade with most moving pieces that happened to Ethereum I think ever. I I might be wrong. At least in the last seven years that we've been in the space it it feels the largest. Um so for Fusaka if it's the next one if I remember remember correctly uh I think it's more throughput related stuff if I recall correctly I can't I can't remember the road by cart >> um it was that but I also think it's more interesting what people are going to do now with backdra because if it is the largest one let's see how it how it pans out and I am evading the question because I can't remember the road map fully by heart but the the long-term one I can maybe speak to that the The idea is to maybe switch Ethereum to risk a risk 5 virtual machine which is also interesting like all of the stuff that we're shipping right now is in the context of the EVM. So how will these transactions behave on risk 5? I think that's a bunch of stuff that people uh writing about this on Twitter are not actually thinking about but just like farming for likes. And then also uh other than that it's like a huge amount of blob space in like the 10x in throughput. So, we're going to see if uh people are actually going to move back from L2s to an L1, which I don't think will happen, but that's one side effect that can happen as well. I don't >> this point. I'm kind of excited about the new ecosystem player coming. Obviously, our good friends at Mega, it's a very interesting one. And I think uh instant finality will be like a huge huge uh kind of part of the future and also the chain like monad and there's like a very few interesting project that are coming up and for me it's like I'm really interested to see like how they will coexist with the with mainet and how like the entire ecosystem is going to shape general. >> Yeah. Yeah. I mean we are also like siloing oursel right. I mean back in Dubai with the the base camp of SUI like people will say oh I just need like to be in this ecosystem and building on top of that. I'm just afraid that we are we're getting a lot around kind of like siling each other because of of course like um you know chain abstraction is is hard you see efforts by by Leifi or LiFi War you know everybody's kind of bring things together but I think on the user perspective you don't want to kind of interact on different kind of ecosystem. I was impressed. I definitely opened my mind a lot two weeks ago in Solan accelerate when I announced the the SVM changes reflecting completely turbine and they did that the efforts will be taking place around 6 to 8 months process. I was like wo how long it takes like kind of like the merge to actually be accessed and got kind of like community feedback right do you feel like the speed of change is uh is quick enough in the Ethereum ecosystem right now? Uh are will users be expecting uh and will be happy with this kind of like change speed? Um or you think like we will start to kind of like get more siloed community uh until maybe we will start to build really chain abstraction and better kind of like uh visibility around the ecosystem. >> I I'm going to say the same thing I said last last week in Prague. Uh I think the slowness is a feature not a bug. Uh I like the slowness the I'm sorry if anybody Cartik was there so he will know. I'm sorry for saying the same thing again. Uh but uh it's very rare for a system to work for 10 years without any any downtime. Uh I think we should keep that. I honestly don't think TPS is what's keeping retail from coming to us. Uh people say I remember when the NFT craze was was huge on Ethereum or I don't know when the things were happening in Ethereum and it like the gas prices were so high and like retail didn't come like the gas prices were high because retail came. I mean they didn't spike just for shits and giggles. It's like what what happened because of demand and I think inevitably if retail comes no matter how much you scale the chain we will use up that bandwidth in some some way shape or form. Uh I think honestly that long-term I I am aligned with the long-term Ethereum road map of uh the base Ethereum stopping to be almost fully for any application and just becoming like a base layer for roll-ups. I think it's interesting that Polado did something similar years ago but unpopular opinion of mine but anyway when we look at it that way I think roll-ups are inevitable. I think roll-ups are the place to experiment and that we should keep Ethereum stable. That's the main thing. And then use roll-ups for interesting experiments like circuit or interesting experiments like I don't know uh stylus from arbitum orbit or from arbitum and all of the others. So yeah, that's at least my opinion. >> I do also agree that roll up are inevitable even though I kind of see there's going to be a big clean up at the moment with like $100 layer twos which probably most of them have zero users or close to zero. So I think in the next few years we kind of will see a consolidation in the L2 space where like the top maybe 10 20 project here to stay and most of the other project they will either pivot or they will disappear completely >> and I think that's true for yeah sorry finish. >> No I was going to say it's going to is similar to what happened with the wallet space in 2017 that everyone was kind of launching a wallet and then we just have now those kind of big players and a few smaller one that are still around. I agree the the thing when when I wanted to jump in I think it's true for ecosystem chains. I think it's unrealistic that we will live in a world where we have hundreds of ecosystem chains. Uh for the rollup future I I think app chains are going to be a thing. I honestly the example I give is uh I really don't need to share the same bandwidth like me buying an NFT ticket to ETH Belgrade doesn't need to share the same bandwidth and same database as Black Rockck doing high frequency trading or whatever they do I don't know buy weapons but basically I don't think that we don't u we have to transact in the same space so this is why I think app chains are inevitable like my star Starbucks loyalty points why should they be on Ethereum mainten or monitor. But no, no, it's it's true. Like it it really doesn't doesn't matter like >> so you say it's going to be very customized kind of future where >> Yeah. Yeah. I honestly I mean I hope kind of depends on it but joking aside never in the history of anything ever has it worked that we cram everything into one place in any industry. I know this sounds like an ambiguous thing but uh if you look at we have multiple cloud providers and then everybody's spinning up their own servers. That can be an analogy for like L1, L2 or like L2, L3. Uh any any any piece of tech ever has developed like this. I think it's foolish to think that we're going to do it otherwise like any other way. >> Yeah, we already saw it a lot, right? Uh also people want personalized experience, right? You see that like people always assuming web 3 maybe people will not do like SSO or or social login, but then embedded wallet boom they're kind of like booming right now, right? And I mean it's it's just different journeys with different needs. And we saw that a lot with naps, right? that people started to have oh security heavy folks who are actually testing out all the security snaps tenderly included right and then you also saw like oh D5 heavy users want to actually have more uh more snaps specific to that so I think like customized experience is something that we need to kind of like satisfy and uh that's probably the the reason also why we're here we're having more fragmentation because everybody's kind of like focusing trying to find different kind of like uh value prop for their uh for the tech technology right um Yeah, but we still need to solve like you know uh stuff what's around you know around layer two fragmentation uh stuff around siloing yourself between different uh uh node EVM and EVM ecosystem because at the end everybody will be uh will be seen as part of like one big package and people coming uh and I saw that I opened really my eyes around the agents uh building and people were saying oh we like oursel our frameworks liza was amazing right so people were starting MVP AI agents And um and then actually like web two falls coming building with Eliza they were actually like after the MV MVP was done they were actually going back from scr scratch and building that from from scratch basically and I was like why you're doing that hey because I didn't get enough flexibility >> so many kind of like we are also like somehow also living in in our uh our own kind of like ecosystem chamber a little bit and uh and some are also like stubborn to how we want to welcome more web two developers in the space and and any user out there. >> But I think everyone is realizing that also the future is inevitably it's also crosschain and also outside EBM even Metam Mask. I mean how you guys were very reluctant for like seven eight years to integrate you know an EBM chain and look what happened next last week with the Solana integration right >> yes >> more more to come yeah the BTCI there is a lot of like different uh things coming but how you will solve like for example layer 2 fragmentation >> but again I think it will solve itself by companies running out of VC money so that's one piece to no I mean that's one piece the second piece is uh chain abstraction will happen in some way, shape or form. So for those that don't know, chain abstraction is the idea that you don't particularly care on which chain you have funds. It will just magically move around. Uh we have it's interesting that chain abstraction is moving slow enough that we have specific companies solving it for for themselves like Tet with USDT0 where the idea is you can spend Tetar or on any chain without needing to think about bridging or anything. You just spend it wherever but it's like a Teter specific or USDT specific thing. Um and then yeah so chain abstraction and then finally I think interoperability stuff. We're going to see how the optimism interrupt stuff happens this year. I think that will be an interesting experiment to see. Uh we have the Cosmos ecosystem where IVC I think solved this already to be completely honest. We might as Ethereum adopt something something similar there. And yeah I I think it will be solved. Uh, I honestly think that a bunch of these things is an Ethereum EOS EVM ecosystem or however you want to call it. We're trying to solve some problems like six steps ahead. At least my brain doesn't vibe with we need quantum compute safe snarks. >> Oh yeah, of course. >> But like nobody's >> doing anything other than D5 like I don't know but there's I hope enough of us in the in the space to work on all of these different issues. I was recently I know I was in Prague like in this panel with DC builder and starting to speak about like from planks to snarks uh and and referring like so many different type of like zk privacies and I was like yeah that's interesting but uh do actually like dubs developer need to know about all this world right now like just refer me like which type of use case do I need to implement which kind of like proofs I think that's for example like walo hanging fruit right and was very interesting because for example on uh on other ecosystem people are very aware that dabs developer are the main audience you want to attract in the developer and builder space and when you're speaking about protocol developer awareness uh it's on the on the other side yes you we can have this discussion but this need to be on a close group right and I think that's one kind of like main challenge sometimes that I have is that 90% of the discussion are on the protocol level instead actually how do you spin up the apps uh what kind of like CLI do we want to actually discuss so at the end we're is going back to the dev side. So that's that's an interesting discussion. >> But also the problem is like I think on the developer side is like it has been historically really hard to monetize developer tooling, right? See what happened with the with things like Infura or there is I mean truffle like there's been many cases where like we had this amazing project that served the community so well but that it was really hard to monetize them. >> Yeah. I mean but developer tooling has also like a life cycle, right? I think uh at the end you're uh you're starting to build a tool that could be actually like for uh for mostly awareness for the builder out there and if you want to commercialize it you have always to find ways right and you see that also with with linear there were 120 people R&D team in consenses for like four or five years and uh you know was supposed to be like almost like could be even like a sunos kind of like department but at the end you see like what R&D can spin off you just need to have the uh initiative to saying okay let's commercialize some product after that but it's a business decision right I I never thought that ganache was supposed to be monetized like that was testing frameworks you see that right >> but truffle teams was a good idea we were scared shitless is tenderly when truffle teams came out no I'm not joking I'm we just released forks our brand wasn't known >> uh and truffle teams came out which I'm not stating that it was a copy I just think it was like similar similar idea at a similar time and it couldn't be copied because it was like two weeks difference or three. It was something. We're like, "Fuck, everybody knows what truffle is." Like we lost, but but it's over. No, no, no. We were We were because it was similar. It was like the truffle teams and truffle sandbox. The ganache sandbox or whatever the name was of the thing was very early similar to what we wanted to do with forks, but we didn't have the distribution and everything else. Um, and it's like different ways to do to do company level stuff. And and I agree like developer tooling is very hard to monetize. I I also think like for the account abstraction stuff we will see if these also bundlers and every else are a viable business model. When you look at Dune dashboards you see a lot of them are not generating actually that much revenue. Um so it is interesting monetizing developer tooling as a topic. I mean this is my day-to-day work at Tenderly. Uh it is hard. It is doable but you need to go in this space as a company or as a team or as a foundation or whichever special little snowflake name you you like for a group of people. You need to go horizontal much sooner than maybe in other industries because it's as a space we progress extremely fast to a very huge level of depth and then we stop which is not normal for any industry and then you need to figure out each time like okay what's the next horizontal thing that's happening and we need to keep pace there and then this is enough and then let's do the next thing and then keep pace and then that's enough. So yeah it's that's definitely definitely hard to pull out pull out correctly. How do you guys prioritize on the developer tooling side on what to build next? Is it like a business decision in terms of revenue? >> Serving developers is a very interesting thing although I guess devil also you know kind of but uh developers are very loud and they yell at you. So I think it's just listening to what people ask for and then the main job of us as tenderly is not coming up with new ideas. Our main job is listening to what people want and then like people say seven different symptoms and you figure out what's the actual cause of that symptom and that's what we ship as a as a product like I can probably all of the products that we ship not by being asked to ship them but us thinking of them we needed to sunset and all of the products where we actually listen to people yelling at us but not like really yelling are the ones that are successful like virtual test is a good example transaction action intelligence. The security scanning scanning piece is literally just everybody coming to us and like you're doing simulations, why don't you do scanning? Like it's not us waking up and saying we should scan these. It's like people yelling at us, why don't you scan these? So I think that's the correct way to >> Yeah. At the end is all about kind of like community feedbacks, right? >> Yeah. And I think I really like what you say about distribution because I think like somehow also like we're building uh and we're thinking like you know technology is what make PMFs, right? But at the end also like distribution and ecosystem building community development community ecosystem building is super important right obviously we're building we're servicing like a community and uh we're trying to do it on a web three way right and cutting corners taking longer I think is also making more sense if you're actually like loyal to to the values but a lot of time obviously like if I'm putting myself in a web two kind of like user shoes I'm saying okay yes you're loyal but at the end I still running decentralized AI in bunch of like silos the AWS kind of like servers. >> So yeah, it is weird. >> It's also hard because when you're running a company, you're starting to lose the company in the community aspect. uh Nenn who who announced us I actually had a lot of conversations with him around this and one of my uh one of my statements is that us as tenderly we cannot have a community we can have a fan base because to have a community you need to have someone to either hold a part of that community in our space that's tokens but then we'll never have a token that's one piece or another that they can somehow contribute so for example I can uh quite sure be that people feel like they're part of the foundry community like we literally have the rust summit tomorrow where a community is going to show up even though like foundaries um is uh front fronted by fronted paid for by paradigm. So even though it's like a very capitalistic entity behind it, it's still a community because people can can contribute. While for tenderly you cannot hold a part of it and you cannot contribute to it. So this means that you can only get a shirt over there and um and be a fan not a community member. And I think that's a hard thing in space like playing this line fully and yeah >> it's a BD work rather than >> I mean it's BD work. Yes. The the way that they say it uh we're more like Taylor Swift less like less >> Yeah. less like a church but yeah >> makes sense. Yeah. Um we don't really have much time but also want to recommend to check the the orbit circle right about community building. It's really interesting to see like intrinsic and extrinsic kind of like incentives and right trying to understand because you know building communities is so hard and people always underestimate the fact that you know product is everything and uh deployment distribution is is easy right especially on the who to hire professional level >> cool do we have question here in the audience I think we're we're at time here I don't know if we have a mic around there you are yes here for unknown Jordan in the audience the OG OG Jordan. >> I back to the uh 7702. Um if you look at Solana or even Algorand or Aptos, you have ability to to sign once to execute a bunch of instructions. It's like executing for an un an undeployed contract once and then it disappears. Now this the spectra is like unix set UID forever and it's kind of scary. Do you know why they went for that much as opposed to just doing the optos or salana sign multiple instructions that that'll do it? >> I think we will inevitably end up using session keys much more. So session keys prevent give you the constraints that you that you mentioned. uh and it's also until the end of the transaction. Uh and then you have the curation piece that MetaMask is doing. So it's not like um so some of the stuff is you need to sign an authorization list which is basically you sign that during this transaction I'm okay to be impersonated by one of these contracts and you actually sign for address for each address of this particular contract and it's only valid for that particular nons. So it's not like you're signing it until the end of time. Now if you do session keys that's a bit different but in the session keys themselves you can um you can constraint it in such a way that it's not until the end of time but it's actually uh up until a certain point or again the the interesting example for me is like twice weekly which wasn't possible before but I think I mean what Franchesco said at a certain point I think the security implications are also going to be interesting because somebody's going to find a way to do an infinite infinite signature or whatever. Uh your question specifically was why did people pick to do it that way? Uh I don't know. I just gave you my interpretation of how it's going to tackle that particular choice. I don't know if you guys have >> looks scary for a naive user, but >> sorry, >> it looks scary for a naive user. They have to know what they're signing. But the fact I also would argue that you're not an uh um normal user by the fact that you could answer ask this question. I think and I mean this is a compliment by the way. I think that you understand it from a technical level and that's why you find it scary. I think unfortunately most people are not going to understand the complexity. They're going to go to the unis swap interface and for the first time they're going to go ah that was this was one transaction. It wasn't two transactions fun and never think about it ever again. So yeah. >> Thanks. Any other questions so far? We're good to go. Okay, sounds good. Thanks so much for your time. Thanks so much panelist. Please, a quick round of applause and a huge round of applause for our host as well.