Web3 Galaxy Brain 🌌🧠

Nicholas: Welcome to Web3 Galaxy Brain. My name is Nicholas. Each week I sit down with some of the brightest people building Web3 to talk about what they're working on right now. My guest today is Reservoir founder Peter Watts. Reservoir is an NFT API provider that makes it easy to build cross-chain NFT interactivity into any app. On this episode, Peter shares how he got his start building music and streaming apps and how music NFTs led him to starting Reservoir with a co-founder he had never met in person. We discuss Reservoir's demand-driven approach to development prioritization, how they deal with tech debt, and why it is that no one else is making a unified read-write API for cross-chain NFT markets yet. It was fun learning from Peter how Reservoir manages to develop so much tooling across so many dApps and chains so quickly. As always, this show is provided for entertainment and education purposes only and does not constitute advice or any form of endorsement or suggestion. Crypto is risky and you alone are responsible for doing your research and making your own decisions. And now, I hope you enjoy the show. So Peter, welcome. Thank you so much for coming on the show today. So to start off, I want to know what is your background? What were you doing before crypto, before Reservoir?

Peter Watts: Cool. Yeah, well, yeah, my origin story is definitely around music. That's how I learned to code. One day, one of my older brother's friends gave us a big case of CDs and I just started burning them. And from there, I got into starting a music blog, which actually ended up on the hype machine. So then in high school, I had this cool music blog. And from there, yes, I started wanting to figure out how to edit the sidebar on my blogger. Started building various apps around music, like a gig booking software, poster printing for bands. I was in a band, but more interested in software around that. Eventually did a music recommendation app. It was called Swarm FM. And so I came to the US, raised a little bit of money for that. Did that for a bit over a year. That didn't work out. I ended up at a live streaming app called YouNow. So that was through some connections at USV. That was crazy because live streaming was, at the time, Periscope didn't exist. YouTube didn't have live streaming. Facebook didn't have live streaming. So all that crazy live streaming stuff happened. And then yeah, this company ended up doing a token. I was trying to do the ownership economy of like, okay, how can you incentivize users maybe of apps, especially power users, streamers that would just go off to any platform. And that was kind of on the side. I was starting to follow some crypto stuff, but got thrown deep into it. The company I was working at decided to do a token. It was 2017. It was the thing to do. And that's kind of, yeah, that was my gateway deep into crypto. But definitely music is my first love.

Nicholas: Crazy. So YouNow did like a ICO on Ethereum?

Peter Watts: Yeah, it was called Props. They tried to do the right thing. And they did like the reggae. So they got it registered as a security. And yeah, it was a terrible, you know, there's so much paperwork. It was like going public, basically. Every feature we wanted to add, it was like, oh, we have to go ask the SEC for permission. And so yeah, we saw far too much about all the kind of legal side of how he tests and all the rest of it. And yeah, the reality was that there just wasn't a path out of that. But the idea that people thought in 2018 was like, okay, early in a project, you acknowledge that it's a security, and then you decentralize and you can kind of walk away from reporting to the SEC. But for the SEC, it just wasn't actually something they comprehended of like, oh, well, who's the issuer now? You know, you can't walk away. So anyway, it was a very interesting experiment. And kind of learn a lot on the legal side of things. But by the way, also learn a lot. I think the idea of the idea was that every app would kind of have a token that like, kind of was a reward, a loyalty token. And actually, this is what led me to NFTs where having every musician or every app or every creator create their own social token or community token was the thing at the time, actually is a pretty bad idea. And does look a lot like a security. And kind of realizing that NFTs are actually the better form for creators to kind of create goods that the community can collect and can potentially accrue value alongside that creator in a way that isn't just straight up issuing a currency. Like you issue a currency, you're kind of stuck with that forever. versus you issue NFTs, they can just be digital goods. Musician could create 10 of them and that's it. They're just like some cool items that people can collect. versus you create a currency and that's just following you for the rest of your life and it can be infinitely divided. So it's just like it's not actually scarce. And so yeah, that was the NFT rabbit hole for me was trying to make social tokens work and realizing that NFTs are the much better form.

Nicholas: And have you taken a look at music again, in particular, in this NFT lens?

Peter Watts: Yeah, I mean, we work pretty closely with the sound team. They're doing some cool stuff. Anthony from The Hype Machine back in the blogging days now works at Zorin and is kind of playing with some cool stuff. And it's one of the use cases I go back to because all my friends are musicians and it's like, okay, what are use cases for crypto? And funnily enough, all of them are pretty suspicious of crypto. But at the same time, they're like complaining about streaming services and how they can't make money. And that's where I think there's interesting opportunities around capturing more value from those true fans. And one of the examples I give is you take a t-shirt. When I was following bands, you might buy a t-shirt. Why are you buying that? You're buying it a little bit to support them. You're buying it to reflect your own social identity. And you're also buying it maybe a little bit of it is also proving you were early. You've got some early version of a band t-shirt or you see people with the tour t-shirts. And we're not there yet. But NFTs can kind of A, be a digital version of that. that's better in many ways in terms of the cost to produce a t-shirt. So much of the money, you're not really supporting the artist. It's so much goes into the cost of producing that t-shirt. You know, A. B, you could be sharing more of a digital identity if you're showing it off. And C, the proof of collecting early is so much better. I remember Last.fm, one of the hacks that I built back when Last.fm was big was proving who listened to an artist first. And then yeah, the last thing with potentially music NFTs is because these digital items, you take a t-shirt, there's no proof of, oh, there was only 20 of these issued. Whereas with an NFT, it's a lot clearer. They don't decay in value. So the provenance is really strong. So maybe I'm willing to spend more on it rather than $20. I'm willing to spend $50 knowing that, oh, it's a bit of a, let me roll the dice on this. Maybe if this artist is big, this will become a coveted item. And so it just feels a lot more efficient and actually potentially effective way for artists to fund themselves. So yeah, it's definitely in the back of my mind. And some of the stuff that Sound is doing and others is interesting. Still early days.

Nicholas: Yeah, definitely. I mean, I think shades of what you're saying are present in PoApp, the open editions, metas of the last year or so. Some of this more participatory memorabilia hasn't gone up in terms of price. I think, you know, free is still the best price. But definitely those things as on-chain history that can be used for airdrops and other things in the future and Guild, etc. Yeah, it makes a lot of sense. It'll be interesting to see if they get packaged. Actually, I might as well ask you this now. Do you think that NFTs are something that the user should know about or not?

Peter Watts: As in like the word NFT or, you know, like a digital item? Yeah.

Nicholas: I mean, you're a little bit familiar with the Reddit polygon NFTs? Like they basically productized their, like a kind of CSGO skins style market of avatars whose artists are sourced from the subreddits. So they had a prior effort that didn't work where they just, Reddit itself sold NFTs. And I think people didn't love it. But then when they kind of turned it around and made it about emphasizing the actual users and allowing them to sell goods to their own communities, it seems to have worked a lot better. But you can see that they're NFTs if you really dig deep. But like the interface doesn't really put that front and center at all. And the wallet is, you know, intrinsic to the Reddit app.

Peter Watts: Yeah, I think it's really, I think it's a great case study. And so like, I guess two things. I think that it's yeah, isn't pushing it that they're NFTs. There has to be some like legitimate reason why that community is interested in these items. But I would also argue in the case of Reddit, the fact that they were NFTs under the hood was this powerful thing. And the fact that they put it on Polygon, rather than putting it on some like, of their own chain that had, you know, none of the existing NFT community, the fact that they put it on Polygon meant that the existing kind of NFT speculating community could speculate on it. And then all of a sudden, the price of these things are going up. I think that created a bit of a feedback loop. And I actually do think like, these things are healthy, where you have like a legitimate, genuine community creating some interesting items. Rather than that being kind of in a vacuum, it's connected to, you know, the composable liquidity with other markets, these items can trade anywhere. That kind of. then all of a sudden, there's people like, oh, this thing that I got, oh, it's an NFT and I see it in a positive light because it's this thing that there's a dollar value attached to it. You know, there's all the negative around money and NFT using crypto. But like, at the end of the day, like that is like an appealing factor. Like I also used to collect coins as a kid, you know, they looked cool. But like, let's not lie, the fact that they, you know, retain value and could go up in value was a feature that was pretty cool. And so yeah, I think the Reddit playbook of like, nicely balancing that where it does have legitimate use in the community, but it is an NFT under the hood. It's like a pretty good model, I think.

Nicholas: Yeah, for sure. So who are your co-founders at Reservoir?

Peter Watts: Yeah, so the company was started by myself in Australia and George in Romania. So and at the time, we actually hadn't met in person. I hired George kind of at the previous company. He's like a, he was I think, 22 at the time, you know, just applied through some random cryptocurrency jobs board, and then turned out to be this amazing kind of technical person who you could throw anything at him. And it was actually a pretty funny story, I think, where the fact that we were on opposite ends of the earth, you know, that can be terrible sometimes where, you know, if there's miscommunication, you know, the whole day is wasted because you don't know what's going on. But actually, you know, George and I, I kind of describe it as like, we're almost like combined to be one person where, you know, I have some idea of like, what we want to explore. We overlap for like an hour, you know, my end of my day started his day, I'll kind of just roughly explain what I had in mind. I go to sleep, when I wake up, he's like perfectly interpreted what I had in mind, built it out, figured out, you know, what's kind of possible. So then I start my day and I can like pick it up from there. So it's almost like we're able to run 24 sevens. It's a pretty eye opening experience of like, the pros and cons of remote. It was, you know, I've worked with people where it's like the complete opposite, you know, the whole day gets wasted because, you know, they didn't quite understand what you meant. So yeah, it was the two of us that started it. And kind of from there, we brought in some people that I worked with in the past, we actually have acquired two companies along the way, which is pretty weird. I think not that common. One of them was like a couple of guys that were building one of the first teams building on top of Reservoir. And so we had this really close relationship and it kind of got to the point of like, now they're trying to start their own company and build things around Reservoir. And I was like, you should just join us. And then later, we acquired another team that was like, again, building something similar. And it was like another young team and they were bootstrapped. And it was just like, we should just join forces. So yeah, that's kind of our story. And we're about 20 people now.

Nicholas: Wow, that's great. Yeah, I remember Rem Koolhaas talks about this like 24 hour architecture studio that works in a similar way by having an office. I think for him, it was like London and Shanghai or something like that. So yeah, I've heard people present this idea, but I think it is hard to pull off. I guess it requires compatible personalities and skill sets and things. So that's cool to hear. So Reservoir offers like so many services and tools, just so many things. How do you group them mentally for yourself?

Peter Watts: Yeah, it is a bit of a challenge of how to kind of describe it all. I think one of the reasons, yeah, so very roughly what we do, we're kind of like describing it now as like NFT trading infrastructure. And I think, really this focus on making it really easy for any app to build buying and selling into the app. And so we're kind of doing that by aggregating every listing, every bid across all marketplaces. We're understanding how to execute fills on any of those protocols. We're understanding how to talk to OpenSea's API and Blur's and allow people to post to those marketplaces or view things from the marketplaces. And really, there's a few buckets. There is people building full marketplaces. And we kind of enable that because they can just take off the shelf, being able to use something like Seaport. And instead of having to understand the complexities of Seaport or even deal with the kind of new versions of Seaport coming out, they can just call a very simple abstracted API that we have. that's like a list API. And you just say, I want to list this token for this price. And we'll just return the exact kind of data blob that you need to sign in order to list it. So we have that use case of people building full marketplaces. We have a really interesting thing, which is the thesis of the company is that NFT trading and liquidity will be everywhere. So rather than just going to a marketplace and that's where you trade in your wallet, there'll be built in kind of, and we have this with a few wallets today, where you're looking at the NFTs in your wallet and you can just press sell. And it's going to go and we're going to find the best quote to instantly sell that whatever the marketplace is. And you can just kind of sell the NFT from your wallet. Or another use case is minting platforms or social platforms that want to show an NFT. Usually when it's minting, that's fairly easy. There's one place that you go to mint it. But as soon as it mints out, then all of a sudden it's trading everywhere. And so we will aggregate that and allow these platforms to kind of, they turn the mint button into a collect button. And it's again, for the user, they can just come one click, oh yes, I want to collect this and they're not really thinking, where am I collecting it from? Who am I collecting it from? It just is kind of one click buying. And it's pretty cool to see some of the social starting to happen. It feels like there's a little bit of momentum going on there and we're able to power. So rather than just be read-only, I'm able to read activity of what people are doing. I think the big focus for us is the right functionality, allowing you to take actions like buy something or sell it or list it, even transfer in the increasingly multi-chain world being at a bridge, all of these complicated actions where there's under the hood many ways to do them. We want to abstract it as much as possible so that every app can have these things as table stakes and drop them in. And I think one of the reasons it seems like we do so much is, funnily enough, in order to build a really good write experience, where you're able to, for example, buy any NFT, we need to aggregate all the marketplaces and have all of the good listing data. But we also need to index on chain and know who owns every token. And so we have all of this really good data in order to have a really accurate order book so that we can then do the simple action that looks like the tip of the iceberg of like, oh yeah, I want to buy the latest token. So much is going on under the hood. And then because we have all this good data, we've started to expose it. So we have an API to get the data. We push our data on to Dune Analytics so that people can query it. We even have an oracle because it was like, wow, we've got this data. Why don't we just... If someone wants a signed message of what the current floor price is, that was just like a small add on because we already had the really accurate data. And so yeah, it can seem like there's a lot going on because there is a lot on the periphery. But really, I think the core focus is around allowing any app to have this interactivity without having to go down the rabbit hole of understanding the functions and all the different marketplaces and protocols.

Nicholas: Got it. So the thesis is NFTs will be everywhere and trading is one of the primary functionalities of NFTs. So NFT trading infrastructure is really just making dev easy, focused on dev UX so that people can integrate this cross-chain NFT trading infrastructure and all the peripheral stuff just happens as a result of building that out.

Peter Watts: Right. Yeah, yeah. It's funny, trading has connotations in the NFT community. It feels more like pro-trading. And so I'm not sure if that's the perfect word. We'll see if we iterate on it. But yeah, just being able to buy an NFT, I do think that will be everywhere. You look at the example of Instagram where they added commerce into the feed. You see a picture of someone's wearing something. Oh, I just want to buy that. And NFTs are perfectly suited for that because they run on these standards and the liquidity can be composable. You can execute it from anywhere. The problem is, there's many standards. So it is open and standardized and anyone can execute a transaction from anywhere. So it's nice and open versus closed proprietary APIs. But then you have this problem of there's actually 10 different protocols that you might need to do and they upgrade every now and then. Right. And so the meme of like, oh, there's 10 standards, we need to make another standard. But that is definitely the mantra in the company is just like, what can we abstract? And so we try and boil that down. Seaport and Blur looks rare and rarible. They've all got their own protocols. But at the end of the day, it's the same action. You're listing a token for sale, you're setting a price, you're setting an expiry. And so we allow a dev to, instead of having to build directly on each of those, they can build on our abstraction. And so that means when new protocols come out, they get them for free. When existing protocols upgrade, we upgrade it under the hood. And they can just really focus on the last mile of what novel user experience they're building. And then, yeah, the last thing I'll say is, I do think trading is a pretty core part to NFTs. Where, like I mentioned earlier, the fact that they can retain value makes them more appealing. So they're not just digital items. I mean, there's certainly use cases where the value of them and trading of them is less important. It's really just about provenance, like pull-ups. But I do think...

Nicholas: Being able to send them for very little expense across the world without losing any risk to their integrity is pretty incredible property. So trading writ large, not necessarily like financialized trading, is obviously intrinsic to their value. So I'm curious, two things. First of all, so you focus exclusively on NFTs or does getting into that peripheral stuff mean that like, well, now we actually need APIs for non-NFT related stuff? Because devs, you know, like to what... Is the expectation that a dev, like, I don't know, building something from a wallet is going to only use Reservoir one day? Or will they all still be relying on things like Infura or Alchemy or QuickNode or Anchor or others?

Peter Watts: Yeah, it's going to be interesting to see how it plays out. Certainly NFTs are our focus. And you know, in part because we just see a really big opportunity there of mainstream consumer applications. And also there's a lot of people, you know, doing or having great products around kind of swapping tokens and bridging and all the rest of it. But at the same time, if someone wants to build an NFT app, and increasingly maybe that's going to be a multi-currency app or a multi-chain app, they might, you know, need something as simple as like, I want to know what balance, what currencies the user owns so that I can show those to the user, purchase something. Or likewise, I want to know the identity of the user. Like who is this person? So what's their ENS name? And so we want to build a really nice identity API that maybe aggregates ENS and Lens and even off-chain OpenSea username so that you can just very easily show a username. So arguably that's not an NFT feature, but it's a feature that an NFT app wants. And likewise, an NFT app wants to be able to show balances. And so I think we're going to build all the stuff that kind of is peripheral to people building NFT apps. And then yeah, there is like low-level kind of RPC and Fiora and the rest of them. It's kind of very different to building on top of Reservoir where it is more proprietary. But at the same time, you're getting things pre-indexed and cleaned. And yeah, I mean, I think there'll always be uses for both. But I think especially once you go multi-chain, like for example, it's not that hard to check a user's balances on one chain. But if you're now on 10 chains and you allow them to easily buy NFTs with whatever balance they have on any chain, the kind of advantage of using...

Nicholas: You're going to start needing to do a lot of the lower level stuff in order to have sort of a coincidence of wants between where my ETH or whatever token is and whatever the contract will accept and wherever it lives. If I want to have that interactivity with NFTs across marketplaces, across chains, I'm going to need to know some stuff about what's going on in all these chains that's not related to NFTs.

Peter Watts: Exactly. And so we actually already today, we allow you to buy an NFT in any currency and we'll swap it under the hood with Uniswap. And so we're already kind of doing... Because it all overlaps. So yeah, that's the story.

Nicholas: Yeah, very cool. So what are the most popular offerings? What's the part of Reservoir that people use the most?

Peter Watts: Yeah, good question. Certainly some of the use cases I just mentioned earlier in terms of wallets wanting to instantly allow you to sell any item in your wallet. Other kind of social and minting platform apps wanting to have kind of a one-click buy. We have people building full end-to-end marketplaces. And so we have not only the APIs for that, but we have an SDK, we have a React UI library, we have an open source marketplace starter that you can just fork and have a marketplace pretty quickly. And we also just have people that come for data. So a lot of the big marketplaces get listings from us, OpenSea listings. And this is something we kind of prioritize pretty early because we don't have all the perfect APIs. And so we made it possible for you to just bulk sync data out of us. because that way, if we don't have the APIs you exactly need to kind of get information on demand, you can at least synchronize the data into your own system and build those APIs yourself. And so yeah, a lot of people are synchronizing data with... We recently launched WebSockets and we have this thing called a sync node, which tries to abstract the complexity of syncing. Whether you want all transfers, listings or bids, you can bulk ingest them out of us. So yeah, that's probably the biggest one.

Nicholas: I'm amazed that you're able to offer all this stuff. It's like a really incredible amount of high quality APIs and even things that are threatening to your... I mean, I guess they're not replicating the software that's indexing, but to sync everything out to somewhere else, even like over a WebSocket, I guess could, you know, I don't know. It's impressive that you're able to offer so many things. I'm curious, how are you able to do so much? Because it's a pretty small team.

Peter Watts: Yeah. Well, it's a team of 20. So it's not that small.

Nicholas: Not that small, I guess. But I kind of expected you to be like a former Facebook VP or something, like with the way that the reservoir is set up. Just in terms of it feeling like this, I guess it is not. But it doesn't feel like it's the first time you've built this.

Peter Watts: Yeah, I mean, some of the team comes from the live streaming app and has some experience. You know, dealing with scale. And our DevOps guy comes from an exchange and dealing with running various crypto nodes. But it's actually for a lot of us, the first developer product. And but at the same time, it feels like a really good founder or team market fit. We've been on the other side of the table. We know how we want the APIs to work. And yes, we prefer to just be at a bulk, sync everything with a WebSocket. And so I feel like we're really good at just, you know, and it's also like a culture that you choose. It's definitely messy, where we have very little roadmaps or forward planning. A, because we just never have time for it. And B, because we just react if someone's asking for something. We try to have a culture of like, okay, is this a fairly small thing for us to do? So even if it derails the bigger thing that we decided to work on, let's just do it because the ROI of making that developer happy. And then what you also find is, we built features for customers and reflecting on it. It's like, some of the pro trading apps, they're not a perfect customer for us because they have all these really demanding needs and they're probably better off building some infrastructure themselves. But we built some features for them. And then all of a sudden, someone else comes along and actually, we're able to use the same features. So some of the bulk syncing of data, even someone we have, Artblocks, for example, uses us for their marketplace, but they want to power the APIs themselves. So when they came along, it's like, oh, yeah, we have everything ready for you to go to kind of extract data out of our system.

Nicholas: Customize to power their site.

Peter Watts: Exactly.

Nicholas: I guess that's down the line on the roadmap to let people just ship whatever plugin to your system so they don't need to self-host a whole separate thing that has to sync separately. Maybe it would make sense. I don't know.

Peter Watts: Yeah. I mean, what's interesting about Reservoir is that it's kind of like a one size fits all. There are some cool developer tools, like even like subgraphs. You think about subgraphs, you can compose your own API and there's cool teams building stuff around that. And then probably the extreme is like talking to an RPC. You can get really custom. Whereas Reservoir is a little bit of a one size fits all. And so there's huge advantages to that. If we have all the APIs you need, you can just build. There's no setup time. You just go straight into building.

Nicholas: And there was a good year or two where everyone was building their own indexer or more, even longer than that, which doesn't seem like really the point of building a wallet app or building an NFT experience. The goal is not to index.

Peter Watts: There's definitely economies of scale as well. And now with multi-chain, I think it's getting harder. But yeah, so I think to answer your question, it comes down to be very comfortable with changing, no forward looking on what we're doing, just very reactive. I know there's some developers that are like, oh, this is our two week sprint and it's sacred. And if you want something new, you can wait until the next sprint and put it on our list. Whereas we kind of have more of a culture of just everything that comes in, you're evaluating and we try also to just knock things out rather than build up a list. And so it's definitely messy.

Nicholas: And yeah, I think just...

Peter Watts: Exactly. And maybe this won't scale beyond a certain point, but I think it works well for us where, like I said, rather than doing all of this thinking of like, oh, is this really the right... You look at our APIs, our docs, we just got a lot of APIs because we just slap new ones on like someone wants it in a slightly different way. We'll slap it on without thinking like, oh, is this cohesive?

Nicholas: Do you worry about the tech debt around that? Or is it like, eventually we'll refactor to some like V2 API. that kind of makes sense of everything that the users wanted that we built along the way?

Peter Watts: Yeah, yeah, I think so. Like, actually, one of the cool things we do is every route is versioned separately. And so what that means is we want to change the response format or the request format. We just like add a new version. So if you look like the buy API is up to version seven, the token API might be on version four. And so like I said, it's like, it's messy. But that's very different to like, oh, we're going to 2.0 the whole API. It kind of is planned and, you know, all done at once. Ours is just like ad hoc chucking on new routes. And there's a little bit of tech debt and like someone gets confused around, oh, wait, there's three different APIs that kind of do the same thing. But I think overall, they're much happier that at least they're happy to know that there's a solution somewhere, even if they maybe can't figure it out themselves, versus like, it's just not possible. So I think it definitely works out. And then yeah, we're actually planning. So everything, most of our APIs are powered directly off our Postgres databases, which is doesn't scale for like really flexible filtering. I want to search this currency between this price range and only from this marketplace. So we're kind of building a new layer on top, like a search layer, basically, using Elasticsearch on top. And so that's kind of an opportunity to almost start fresh. So rather than even needing to refactor what we have, you know, as we layer on new things, we can kind of take old lessons from the past and kind of design it a bit more thoughtfully.

Nicholas: So you were saying about like some of the pro services, like. I know OpenSea Pro is somewhat built on Reservoir. I was curious, why would they use that rather than whatever OpenSea has natively or like, you know, internally? And also like, is it? yeah, you kind of alluded to this, like, maybe a bigger customer like that is harder for you to handle. So is your ideal customer like more individuals or new projects? Or is there a difference in the size that you're targeting? Or yeah, how do you relate to these like, bigger customers who could afford to build it inside?

Peter Watts: Yeah, I mean, so OpenSea Pro do a lot themselves, because obviously they were, you know, back in the gem days, I think they started a similar time to us, which is coming on close to two years. So they do most of it themselves. And what they use us for is kind of the long tail of protocols. It's like, I think aggregation is becoming more of a commodity. And so for gem to have like the latest protocol that has point 1% liquidity is probably not worth their time kind of going and adding that it's not like that's going to get them users. You even look at blur like that, they don't kind of necessarily bother to get the long tail because it's not as critical. So anyway, someone like OpenSea Pro uses us for the long tail, but they do most of it themselves. My comment about the pro trading tools is more like, they need really kind of real time and some of these other things. And as soon as metadata reveals, it needs to be like in under a second, they've got the new metadata, which increasingly, we want to make sure that we support that as well, because it's not just pro trading tools that maybe want that. But it certainly was more of the... Yeah, I think you've got the spectrum where it's like. if trading is your focus, it might make sense for that for you to build a lot of it in house. But if you're a social app, or even you're a creator that just wants to run your own marketplace, or you're a wallet, the trading is just one feature amongst a lot of things that you're doing. And then it makes a lot of sense to use reservoir and if reservoirs kind of gets you 90-95% of that functionality, that's great. Whereas if you're a really kind of pro trading app that's trying to take on Blur or OpenSea Pro, where don't you just use us off the shelf and expect to beat them? You're gonna have to do some stuff yourself. So yeah. And the other thing I'll say is, I think for NFTs to grow, we do need these mainstream use cases and beyond just pro trading. And so that's kind of where we, in terms of prioritizing features, we're very much interested in empowering that growth and who's building things that can actually bring new users in, new use cases. And so we tend to kind of, we'd love to build real time metadata revealing for someone who wants to build a pro tool, but we do have limited resources. And so it is a little bit of kind of saying no to things and trying to prioritize. We see kind of like, I guess like the vision of NFTs and liquidity being everywhere is more true on the consumer side, maybe than the pro side. Like on the pro side, it almost does make more sense to just go use one tool and everyone's there. Whereas on the consumer side, I think it makes more sense. It'll be much more fragmented. Every app will want to sprinkle in a bit of NFT trading functionality. And so that's kind of where we should focus.

Nicholas: That makes sense. Yeah, I guess. I mean, let's talk about NFTs for a second. So what is an NFT to you? What is an NFT to you? Yeah. How do you think about it?

Peter Watts: Getting very meta. Yeah. I mean, I just think of it as kind of digital items. And I think that's pretty broad. And obviously, I think the scarcity is a really important piece. And so the fact that anyone can create digital items that are provably scarce, unlocks all sorts of interesting things, right? Whether it's an artist creating some collectibles, I'm more like a musician, whether it's art, whether it's tickets, whether it's real estate coming on. I think those things are definitely more in the distance. And I do tend to still kind of think a lot in terms of collectibles.

Nicholas: Yeah, this is my question for you. Is Reservoir, does the premise currently have any allegiance to like NFTs? are art writ large? Or if NFTs are like, I don't know, key fobs for condo buildings, is Reservoir just a setup for that?

Peter Watts: Right. Yeah, well, yes, if you've got NFTs that like, yeah, are just designed to be like held or you know, like, yeah, there's some key access key to something, then I think it's less relevant. I think we are kind of focused on the subset where there is scarcity and potential kind of trading and therefore you also want to know what is the price of this thing? What is the value of this thing? And so there'll be use cases, even PoApps, right? We haven't built anything around PoApps. We probably will because there is that like identity layer to it. And so I think that a lot of apps might want to bring that into the kind of... They might want that in depth.

Nicholas: They might want special affordances for PoApps because of their specific nature.

Peter Watts: Yeah, yeah, yeah. So there is like all sorts of like those experiences where you unlock things based on what you're holding. And I think that's a good thing. But I do think that's not trading per se, but is definitely more relevant. Like I think the common thread is like NFTs with scarcity. So like they're either scarce, which means you want to trade them, they're scarce, they're showing your identity or they're scarce, which is unlocking access versus like if they're just kind of maybe digital, some form of digital item that is like less scarce and just kind of piggybacking the blockchains to kind of record ownership, then I think it's kind of less relevant to us probably.

Nicholas: We were talking a little earlier about like from the music background, thinking about NFTs as collectibles to support and rep, you know, how early you got in. But this kind of proliferation implies like a proliferation of collectible secondary markets on the internet in a way that there's definitely niches of physical objects that behave this way. But it doesn't feel like everyone you know is a collector and trading collectibles yet. However, NFTs kind of, if this is to be a popular format, it seems like the internet would be like taken over by in the big way by trading culture of collectible items that are limited issuance. Do you think that's likely?

Peter Watts: I think the more that I've thought about it, it's like there's explicit collecting and then there's like kind of implicit collecting. A lot of things when you squint at them is some form of collecting. So with music, even just, I used to listen to far more new music. I would always be listening to new music and then I find something that I like and like, cool, let me put on the shelf and keep listening for new music. And that's a form of collecting, trying to just find the good things, but then not even like enjoying them, but just on the hunt for more good things. And same with fashion, where it's like, okay, that is a form of collecting or being early to something. And so I think it might not necessarily look like, I think there's already a lot of behaviors that aren't like explicit collecting and trading, but they can be digital forms of them. And kind of run on NFT rails and benefit from markets being out of form around them. But that doesn't mean that that's like, I don't think it has to become really explicit. And I think it's better if it's not really explicit.

Nicholas: Yeah, that's what's odd about the NFT market up to now. It's like everything is selling itself as a future heirloom collectible item before it has established its cultural relevance. Whereas like, I don't know, buying a t-shirt at a Madonna conference back in the day, it's not as much like this will be worth something in 40 years. So it's an interesting, I guess what we've yet to see very successfully is that people using NFTs for things that are already seriously culturally relevant. I guess that's why I bring up the Reddit example, because there is a legitimate reason you want to look a certain way on Reddit. And then it sort of has this maybe to some extent, incidental collectible value. But maybe the next bull market, we'll see that happen more.

Peter Watts: Yeah, yeah. The Reddit is a good example. But even usernames, right? Like hopefully, there seems like a small chance that maybe someone like Elon Musk will experiment with like dormant usernames, like putting them up as auctions as NFTs. And so yeah, that is, people do collect usernames, they collect domain names.

Nicholas: That's true. There are parts, people definitely collect links and liking is a form of collecting, even if it's centralized and not permanent. So there are these habits of arena, pinboard, delicious, these other things where people are Tumblr even curating collections of things. So there is, you're right, there is a collecting habit. So how do you manage to, from like an infrastructure point of view, as we're going into this multi chain world, how are you actually managing a back end that is able to deal with all of these different blockchains that are popping up all the time and at an increasing pace? How are you able to do that?

Peter Watts: Good question. I think we're just like, rushing head on into it. And we maybe will feel some pain down the line. But at least one interesting thing is that the way we're set up today is we kind of like have infrastructure per chain and it's fairly segregated. And going back to what I was saying previously about this elastic search layer that we might build on top, that can be like a multi chain that can have data from all chains and allow you to easily find things across chains. But most of the things you need are one chain at a time. And so, yeah, we kind of have an infrastructure that is running on Kubernetes and Terraform and all those cool things that we can fairly easily spin up new chains. I think right now we're taking an approach of just like spin them up and figure it out.

Nicholas: Yeah, because you're pretty fast on Zora and I think every time something new happens, you're right there.

Peter Watts: Yeah, well, I mean, it's a decision as well of like, are we going to prioritize that? Because if we don't do that, then if someone wanted to be on Zora and we're not ready, then it's a dead end for them. They can't rely on us. They can't kind of build on top of us. And so we almost want to be faster than RPC nodes. That has been a decision basically of like, we want to be before anyone decides to do that.

Nicholas: Because confidence for, oh, I can pick this provider and then whenever it comes, they'll be there supporting whatever the newest stuff is.

Peter Watts: Right, and the newest stuff is easier. It took us like a month to get Binance chain properly indexed because it's so huge and has all this history. Whereas on day one, Zora's got nothing. So it's like, okay, well, let's just spin it up and be ready. And there's always some kinks to iron out. And like I said, it might not scale in terms of managing them all separate infrastructure. You probably want to run multiple chains on the same database eventually. But for now, we've just kind of kept it all fairly segregated. And we're just kind of iterating on how we can automate it more so that each one requires less babysitting. But yeah, I do think a lot of it just comes down to strategically, you can either focus on the chains where there's activity, or you can kind of take a position of like, we're going to be very aggressive and support them all so that you can safely build on top of us and know that as soon as you feel like you want to go to a chain, we're already there.

Nicholas: It's interesting to BSC example you give because I think a lot of people in a certain part of I don't know what you'd call it, like intellectual CT or something, would not choose to support BSC. But for you, I guess it makes sense because you're trying to be everywhere.

Peter Watts: Yeah, well, there is an exception. We're everywhere EVM. We certainly we toyed with kind of exploring other chains. The beauty is like we have one indexer and one kind of infrastructure that gives us every EVM chain. Whereas if we almost need to build the same...

Nicholas: Doing work.

Peter Watts: Yeah, it's real. You know, we'd almost have to do the same amount of work kind of to build a Solana one and then we get one chain from it. And same with like Tezos and Flow and even Bitcoin ordinals and like, there's cool things happening there. And then not to mention, I think we will start to see like, better cross chain composability in terms of like bridging the UX, some of the things you can do between EVM chains or particularly chains that are on the same like super chain or orbit or these kind of subnets that are emerging. So that's how we're going to do Avalanche as well. I don't know if that's out yet, but it's like if it's EVM, cool, we're in. And maybe one day we'll do non-EVM.

Nicholas: Yeah, I imagine it would make more sense to just acquire some small API provider doing ordinals for third party developers or something and then rather than build it all in-house.

Peter Watts: But like there's people doing... So I think that's the difference with us as well. It's like, okay, so actually just like indexing the data, like maybe not that hard. But then you've got to kind of do the execution, you know, being at a list and bid and buy and sell. And there's different protocols, there's different wallets, there's different wallet software and we're more than just a data API. We have like our React UI components, like getting all of the whole reservoir stack compatible with those chains is like a huge effort. And at least for now, you know, the ROI isn't quite there.

Nicholas: But the dividing line is if it matters to convincing EVM devs, we have everything you need. We have everything, every chain you could possibly want and every NFT interaction you could possibly want will be supported here. Yeah, I guess I was going to ask, this was one of my questions, how do you decide what to support and what to ignore? I guess that's the answer. Do you... I get... Mike Demaris asked this question. I forget how he phrased it, but I think the more polite rephrasing is like, why do you not have any competitors? Maybe you do have competitors. I know Alchemy does a little bit of NFT stuff, but it doesn't feel like what you're saying. Read and write.

Peter Watts: Right. There's a lot of NFT data providers that are doing kind of, you know, there's all the RPC node services have kind of NFT APIs. There's kind of indexing ones like SimpleHash, which does do all the non-EVM chains. We even use SimpleHash kind of under the hood for some of our indexing. Because again, like I said, the pure metadata and read side of things isn't really our focus. And so yeah, there isn't really anyone else doing write in the same way that we are. I think in part because it's a schlep of dealing with all the different protocols and transaction execution. There was one team and we acquired them. That was the module team.

Nicholas: Yeah, I was going to ask about them. They were doing something similar.

Peter Watts: And so they were doing some of the similar stuff. And that's kind of why I was like, we should join forces. And yeah, I mean, I think the market is also small scale. It's not like swapping where there's kind of a few people that, you know, there's the equivalent in the token swapping world where you've got 0x, parent swap and one inch. I think NFT is just way messier. You're having to deal with different tokens. You have to do all this upfront indexing, right? I think with token swapping, you can just like get quotes on demand. Like I'm going to ask Uniswap, I'm going to ask 1inch, I'm just going to like on demand, ask for quotes. Whereas with NFTs, there's so many of them, you kind of need to take a different approach of indexing everything in demand and keeping an order book up to date with every state. Like we do something for the wallets where we'll give you an API that tells you all the tokens in the user's wallet, sorted by which ones have the best top bid. And so like, unlike token swapping, where the user might say, I feel like I'm going to swap some USDC. And so once they've made that decision, then you can call an API and get the price of USDC to ETH. We almost need to... We're influencing the user. They don't know what NFT they want to sell maybe. And so we need to know kind of how much they can get for every NFT in their wallet so that we can show them this sort of view and put at the top, oh, this is the one where there's like actually a good deal for you or send you a push notification because you got a good bid on an NFT. So yeah, I think it's just a very different game. And there is a lot of upfront indexing and costs. And it's pretty messy. And Reservoir is doing it perfectly.

Nicholas: So yeah, I guess I was going to ask this question, like, you know, what does it mean to build an API company on top of freely indexable data? And I guess what you're saying is like the sweat equity of especially solving this for NFTs is enough to really be a moat. It's not so easy to replicate, even though you can say it in a sentence. It's actually quite hard to do all the work. At least that's what it sounds like. Yeah, definitely.

Peter Watts: I mean, it's no different to the fact that, you know, Coinbase and Binance, they're all on open data. The product that Coinbase is selling, you know, you can go get in a number of different places. And like with an API product, it's maybe a little bit different in the sense of your customers or developers who maybe could just go and do the same thing with the raw data. But there's absolutely economies of scale. And yeah, it just shouldn't be a focus for so many people that are building more of those consumer experiences. But yeah, I do think the moat is just like that culture. You know, we definitely look at Stripe a lot as a kind of role model where it's just amazing developer experience and we're just iterating on it so quickly. And we're building a reputation there. And that's the moat more than like, you know, even Stripe. They're not the only ones that can do credit card payments and banks. It's kind of a similar thing where there's some upfront schlepping of building integrations into all of them. But then otherwise, it's mostly just about reputation. And I think what Stripe's done well is like, even as they've grown, they still feel kind of nimble and responsive and kind of building what people need. So you kind of don't feel like you need to go use something else. So yeah, it's less of a data moat, like a state moat where it's like, oh, we own the network. And it is more of a kind of brand moat of like, we can trust these guys. We know they move quickly. We know that they have the features that we want. And then like a little bit of economy of scale moat. Especially once you go multi-chain, it's like, yeah, we don't want to index 20 chains. We just want to make a few calls to someone who does.

Nicholas: That's not where we make our money. That's not the point of what we're doing. Do you pay any attention to what chains are popular? Or do you have a sense of, based on the analytics that you have access to from use of the APIs, do you have any observations you can make about NFTs on L2s? Or how the market has changed over the last 12 months or 24 months?

Peter Watts: Yeah, I mean, it's hard. It's hard to bootstrap an NFT ecosystem. Because if you think about it, it all comes down to the issuers, like the creators.

Nicholas: Really? The big thing was the blur was all about the traders. We really focused on traders.

Peter Watts: Well, no, it was on Ethereum. There was already an ecosystem. I was more meaning, yes, if you're a new chain, and you want to like, get an ecosystem, even if OpenSea goes to a new chain, it doesn't mean anything if there's no good NFTs there to trade. So it's all about having NFTs on that chain. And then if you're an NFT creator, who's building this thing, you're all in. It's huge risk to go onto a new chain versus going to a safe chain where there's an existing community that you can enter. And so yeah, optimism really struggled. You had kind of what was it called? There was a marketplace like trying to make optimism. NFT is a thing. Yeah, you know, and it was kind of tough. Even to this day, it's pretty small. So really, Polygon is the only one that's in my mind, you know, like, I don't too closely follow like the Avalanche and the Binance. I think there is some things going on. But like, Polygon is probably the only one that has like legitimate, like really top tier projects on it. And so it did a really good job of just positioning itself as like the cheaper Ethereum.

Nicholas: Is it mostly OpenSea?

Peter Watts: It's quite possible. Yeah, that the fact that it's been supported by OpenSea for a long time, but I mean, it goes both ways. OpenSea supported it because it had some legitimacy.

Nicholas: But I mean, is the activity there? I mean, obviously, there's minting directly on the contract, but is most of the activity that you perceive there like OpenSea activity? Or are there other things that are popular?

Peter Watts: It is. It is. OpenSea is probably the biggest on Polygon. Yeah, there's a few others. There's someone trying to do like the blur of Polygon. I forget what it's called. But it's also a lot of like project specific. So I think like Decentraland kind of like had always had their own marketplace and same with like Zed Run. So a lot of the early projects on Polygon just did their own marketplaces because they could do maybe like fully on-chain marketplaces and it wasn't too expensive. But I think a lot of it's OpenSea now. But then I'd say like one of the more interesting things is like the Zora chain, which is like I think what it's proven is when I said it's like really hard to bootstrap an ecosystem on a new chain, it does feel like the one entity who can pull it off is a minting platform because they have existing customers that are coming to mint so they can kind of set the defaults and move people onto a new chain. So it's like if you just create your own chain and you're not, you know, so if you're Optimism, they never had leverage to make Optimism and FT as a thing. Whereas Zora has a lot of leverage to make it a thing. So it's been kind of pretty interesting to see that play out. Maybe there's just a bit of like new toy shininess going on. And there's certainly no secondary trading happening on Zora chain yet. It's more just minting.

Nicholas: Although the OpenSea just announced support for it today. We're recording this July 7th, 2023. But you're right. It will be interesting to see if this leads to a divergence in manifold and Zora primary approach. But I agree, it does seem like Zora is very well positioned to pull this off. And the other chains, the other L2s that exist, I mean, maybe the closest thing was like Magic on Arbitrum to like a genuine NFT centric approach. But the L2 infrastructure providers don't have very much NFT DNA or legibility to NFT creators and they don't have collectors either. So it's really hard to boot up those markets. But Zora with its primary front end, successful front end for primaries, can very easily just slot in their own chain. And people, it seems like people are increasingly willing to bridge, which is pretty interesting. It will be. I'm curious to see if on-chain becomes more relevant in that world, like on-chain order books or other on-chain trading mechanisms, rather than the dominance of off-chain order books in the ETH land.

Peter Watts: Possibly. Yeah, I mean, maybe post EIP 4844 or whatever it is when L2s really get cheap, maybe, because they're still not that cheap. And yeah, Magic is a good example of like, I would call Magic an issuer. Like they basically have this launchpad and so they onboarded all of these games and projects to Arbitrum. And so that's kind of, that's definitely where the leverage is to make these ecosystems happen. And maybe we will get to the point where you, because you can just buy on any chain easy, like you might get an issue on optimism or base and someone can just easily buy that from Ethereum. And that if the marketplace is abstract, what chain you're on. So it might be like less of a barrier going forward to launch on an alternative chain.

Nicholas: I'm very curious to see if it's possible. And I don't know the ins and outs of it, but my impression is DYDX on Starkex is some evidence that you can maybe have your order interactions on one chain and the effects happen on another chain. or at least you may be trading symbolic ownership on one L2. that represents ownership of assets on L1 or assets on other L2s. But certainly like being able, you know, I have some ETH on OP and I want to go buy Azora NFT and I don't, the wallet just handles it. That'll make that very easy. But it would be interesting to see if it shakes up marketplaces to be able to have like cross-chain data allowing for the exchange of that. Perhaps the marketplace exists on one app chain, whereas the assets can live on all different chains.

Peter Watts: Yeah, I think it's definitely possible. I mean, like bridging to another chain is just depositing in a contract. And so like, what's the difference between, you know, what you do in Blur where you're often, you know, you deposit some money in a Blur contract and then you trade with it. It's like a similar UX, if you just call it depositing. And then you trade on the L2 versus bridging.

Nicholas: What if a future Seaport version supports, you know, you set approval for, I don't know, you know much better than I do, you set approval for Seaport to be able to move your stuff. and then you're doing your listings on Zorachain and people are executing transactions on Zorachain that are propagating to a Seaport on mainnet that's willing to do it. You've still got the same approvals pattern for the trading permission, but like no vaults required necessarily. But maybe you're actually doing on-chain transactions on some lighter weight, more centralized chain.

Peter Watts: Yeah, yeah. I mean, one problem is it's quick to go from to bridge to an optimistic roll up. It's not quick to bridge back, like especially for NFTs, but maybe there's work around for that.

Nicholas: Yeah, that's a very good one.

Peter Watts: I think there's some interesting things that can get done. You know, at the same time, off-chain is free. It's hard to compete with free. And so even in cheaper execution environments, off-chain order books might still, you know, be preferred.

Nicholas: My crackpot hypothesis is that Zorachain would subsidize bidding and listing on-chain on their own chain when it's really just a SaaS fee that they're already paying, more or less. But I don't know. It's not clear to me if putting it on-chain maybe increases perceived decentralization, but more importantly, composability. I don't know if you get some advantage for being a dev ecosystem where people can build on top of on-chain orders or on-chain marketplaces in a way that they wouldn't have access if it was just like order stored on some server somewhere.

Peter Watts: Yeah, yeah, a little bit. But one of the challenges is like, even if it's on-chain, you still probably need someone to index it and make it easy to use. And so it's like functionally not that different if you're still maybe using Reservoir to kind of access the order book. And then the other thing is, which is a big part of our thesis is like, would that ever get to 100% of liquidity? Or would it maybe like, maybe it gets to 80% and you're going to want that other 20%. Reservoir started as an order book. It was kind of like the idea, but the realization was, we're not going to be able to get 100% of liquidity in it. And someone who's coming wants the best price across the whole market. And so that's why you need to get into aggregation because then you can piece together. Even if one destination has 80%, you can go get the other 20% and then offer someone 100%. And so I think there will be really interesting things like that. But part of me feels like there'll always be a little bit of fragmentation. And that's a good thing. That means there's innovation, people are trying new things and there's competition and new models like bonding curves and pools make sense in some scenarios. So I think it's like the fragmentation is good. And so that's kind of our thesis is embrace the fragmentation and just kind of like, make it easy to deal with rather than trying to get to some utopia where everyone uses the same system.

Nicholas: Yeah, I think there'll always be business for Reservoir. It would be interesting. I mean, I think like PseudoSwap, I had Owen on the show a couple weeks ago and I think PseudoSwap would be such a, the 1155. support for PseudoSwap would be so compatible with Zora 1155s and especially on an L2 where it's cheap. It doesn't really make sense to have individual listings for an NFT with 500,000 editions out there. It might make more sense to have something more like an AMM. But you mentioned, I mean, on the topic of aggregation, I'm curious, like Blur originally didn't share their API, right? You couldn't access their order book originally.

Peter Watts: Yeah. All the marketplaces.

Nicholas: Why they changed their mind?

Peter Watts: I mean, it's an interesting situation. Like everyone, I think there's like, you can't close things down completely. And so, yeah, I mean, I think hopefully where we get to a point is everyone realizes that there's always ways to access it. You know, X2, Y2 tried to block Blur, Blur just found ways around it. So I think we're settling in an equilibrium where, yeah, the liquidity is more or less open, even if you have to jump through some hoops to get it. And Blur is not winning because they locked down their liquidity. They're winning because they have a good UX for pro traders and incentives for pro traders. And so, yeah, I think that is kind of more or less the equilibrium that is playing out.

Nicholas: So the interface and I mean, to some extent, originally it was optional royalties making them and no fees, which I guess, do you know the score on that? I know they were increasing the like by 0.5% the amount that they would honor creator fees some time ago. Do you know if like OpenSea is still charging, does OpenSea charge fees now? And what's the status of honoring creator royalties? Good question.

Peter Watts: OpenSea does have, I think, their fees back on. And maybe with OpenSea Pro, though, you can avoid them and just pay like minimum creator royalties. Yeah, I mean, Blur said that they would increase fees over time, creator royalties over time. But they also said, but if anyone undercuts us, we'll match them. And so I don't think they'll ever go higher.

Nicholas: But they're at like 0.5%, I guess, or that was the first number.

Peter Watts: Right, they've been at 0.5%. But it does feel...

Nicholas: And OpenSea is back to 2.5% or less?

Peter Watts: Yeah, they're back up to 2.5%. And yeah, it's pretty incredible. OpenSea is kind of stickiness with retail users. I mean, it's a similar thing with Zora. You know, I think there is some price insensitivity. People don't mind paying a little bit of a fee, which I think is good. It means that there's businesses that can be built. It's not all going to go to zero. I mean, even Blur, there is a willingness to pay this 0.5% on every trade at least. I'm sure Blur will probably try and transition that from creator royalties into Blur protocol fees. And that's basically what Looks Rare has done, I believe. I think Looks Rare just got rid completely of the creator royalties and it's just a 0.5% to them. And so yeah, it's a bit of a race to the bottom on the pro-trading side of things. But yeah, I mean...

Nicholas: It's a weird equilibrium that we've reached here, or at least moment, we've reached where Blur is honoring 0.5% worth of creator fees, not charging any protocol fees, I guess. And OpenSea is charging protocol fees. And maybe Blur just served to allow them to make 2981 fees optional, I guess.

Peter Watts: Maybe. Yeah.

Nicholas: Like, it didn't really steal the whole market from them in the end. And it seems like as soon as the incentives go away, people lose interest in Blur. And OpenSea is back to charging its fees, except now it charges less creator fees, right?

Peter Watts: Quite possibly. But Blur's incentives will end in about 20 years at the current rate. So no, I'm just making fun of the dragged out seasons with no announced end times. I mean, it's definitely a masterful execution of squeezing every bit out of the incentives. It's pretty crazy to watch.

Nicholas: So we talked about how pro interfaces are. maybe like, maybe they need some of their own special sauce on top because of the kind of latency expectations they have and requirements they have. How are people using Reservoir and other applications? Is it like, I know there was a wave of collections doing their own marketplaces and maybe assessing their own fees or having no fee marketplaces. Is that the most popular other use case? Or you mentioned wallets also?

Peter Watts: Yeah, yeah. So it depends like what vector you define popular. There is a lot of kind of like single creator or community marketplaces. But what's interesting about those is like, that doesn't mean the volume will go there. And so in some ways, it makes sense for every creator to have that. If you can just drop it in, why wouldn't you have a storefront on your main website that offers a better deal? If you're bringing the users, then you can do that. So this is what I often say is like, someone like a Zuki or some of these other... Their users are the NFT, the existing NFT community. So in many ways, like Blur and OpenSea is the vertical marketplace for that user base. So like a Zuki, if they will create their own marketplace, they don't really have their own users that they're bringing. that aren't, who don't know of Blur and OpenSea. Whereas you take like the Reddit case, Reddit's bringing its own audience. And actually, maybe like a better example is Sound.xyz. They very much kind of control a lot of the music NFT collectors. And so when they created their secondary marketplace, they were able to do a lot of the volume on it because they actually kind of controlled the funnel of where new users were coming from and they could just kind of have them use their marketplace. Whereas we often see creators create a marketplace and like no volume will happen on it because their users are quite happy to just keep using Blur and OpenSea. And so anyway, that is like a use case. And a lot of people are doing it. And I think it might drive more volume once more. NFT projects are actually bringing their own audience, not like competing over the existing audience. But then yeah, I think we're more interested and invest more in the kind of like embedding liquidity into various apps. So whether that's wallets, whether it's social apps or minting platforms, being able to embed the secondary liquidity or even being able to embed minting. I don't know if we released it today. I haven't checked my messages since I woke up. But we now have an API that lets you mint any NFT and we'll be watching other people mint it. So like in the same way that we aggregate OpenSea and Blur and Looks Rare on the secondary side, we'll now aggregate Manifold and Zora and Sound and Artlox. There's a lot of them on the minting side. And actually on the minting side, you have lots of custom contracts. So you can't pre-build an integration for them. We have these generic handling where we watch other people mint it and we guess the parameters of which is the quantity parameter or whatever it is and simulate being able to mint it. So that hopefully you can mint as many possible NFTs through our API as possible so that people can build. Imagine a social feed, that activity item that says, Peter minted this NFT. Well now you can just make them have a button there that says, I'll also mint it. So that's the sort of stuff that we want to power.

Nicholas: So basically, the interfaces that are driving the most volume, aside from major platforms that are customers in some sense, but are really, really doing a lot of work on top of how they're using the APIs, the major successes are ones where people like in Sound, for example, people are probably prefer browsing those NFTs on SoundXYZ than on OpenSea where it's really not set up for music. So it makes sense that you go to an artist's profile page, you see their other songs, and then you might want to engage in the secondary for their prior songs, in addition to participating in the primary. And then also you've built this thing where you can now spin up like a Mint.Fun competitor, sort of easily, I guess.

Peter Watts: Yeah, well, I mean, it's similar to OpenSea Pro. Mint.Fun is the best at minting and I think will continue to be the best app for minting. But then there's other apps that may want similar things, but it's not their core focus just to show what are some trending mints, what's going on. Or like a social app, that's not really competing with Mint.Fun. It's not like the core action is, I want to always be minting and finding the best mints. If that's what you want as a user, you probably should go to Mint.Fun. And they've, like we discussed with Pro Trading Tools, they've built a lot of their own infrastructure because it's their core product. They probably should build their own infrastructure rather than rely on us. Although maybe they can use us, we're friends with those guys, and they actually use us for some secondary functionality. Maybe they would use us for some long tail mints that they're not interpreting. that maybe we interpret. But at the same time, if minting is your core focus, you probably should build some of it yourself. Just like if pro trading is your core focus, it might make sense for you to build it. And it's really like the peripheral things where it's like that liquidity everywhere, where it makes more sense to just kind of like drop in reservoirs functionality.

Nicholas: Right, right. Interface, Warpcast, Lens, Lensster, this kind of thing. I'm curious, how do you avoid getting like rugged when you're simulating these transactions? How do you make sure that what you're suggesting is a good thing to mint is actually a good thing to mint?

Peter Watts: Yeah, it's a good question. And yeah, we kind of just, you have to have, like you said, we're kind of like doing this like continuous simulation just to make sure that the equations don't change. And yeah, I think it's an interesting question. Could someone design a contract where it looks like you can mint it, but then after a while...

Nicholas: Maybe I guess the ideal would be like you issue, it depends, I'd have to test it. But like if you're paying attention to events rather than actual state changes, or state changes that are events rather than state changes, like people were spoofing OpenSea for a

Peter Watts: while,

Nicholas: making it look like someone had like a famous artist had minted an NFT and then they went to sell the NFT because OpenSea activity history was paying attention to events, which can be fudged.

Peter Watts: Yeah, that's actually a good point. So when we simulate it, you do get an event saying that there was a transfer. But we actually have like a custom tracer to actually make sure that the state changed. So this has been something like, you know, a lot of the RPC services will allow you to use custom traces. So this is like one of the challenges of expanding to many changes, making sure we can do our kind of custom simulation methods. But I think the other thing I'll say is like, also, usually a lot of these are free mints. So people aren't even like sending money and they're just calling a method. And so like, the risk is usually pretty low. And this is something like a philosophy we have in general, like we are allowing people to take these right actions. And so it does like increase the risk of some of the things that people are doing. But at the same time, you know, A, it's getting a lot easier. And I think a lot of wallets already are or they absolutely should be kind of showing you a preview of like, here's how your balances will change. You know, this will go up and this will go down. Amen.

Nicholas: They definitely should do that. It's not that hard to do that.

Peter Watts: It's really easy to do that. Kind of crazy that that is not insane.

Nicholas: And then the other thing is, yeah, unlike maybe, Maybe that's something they need as a service even actually, because it seems like most of them don't do that. Yeah, yeah.

Peter Watts: We have a little JavaScript snippet that kind of shows how to do it if anyone's any wallets are listening and are interesting. But yeah, so on the one hand, I think wallets will increasingly help to protect users against some of these things. And the other thing is like, we always try and avoid like having approvals or holding any user funds. It's always like a transaction at a time risk profile versus like a honeypot where someone could come and steal everyone's assets. So we try and be like very low touch. We're just kind of like connecting people. So that like, it's just not really the damage you can do is almost like not worth bothering to try and like, yes, there's maybe like loopholes where you could trick someone into minting and then haha, they didn't actually get a token. But like, you know, the damage of that is pretty small.

Nicholas: Right. Steve Klebanoff wanted me to ask you about creative use cases for your NFT price Oracle. Maybe first you could just explain what that is and then how someone might use it.

Peter Watts: Yeah. So like I mentioned, we already have the data around every listing that is currently valid. So we know the full price of every collection. And at that point, a lot of on-chain things. So some of the use cases we've seen is like, you know, lending protocols that might want to know what to liquidate at or at least, you know, I think maybe when you're putting something into the vault, understanding like the rough value of it. And so it's effectively just an API call, but we sign the API call, you know, without Oracle. And so what that means is like an on-chain contract can know that it came from us, someone didn't manipulate it. And just to take it aside, one of the ways that Chainlink Oracles used to work is that everything would be pushed on-chain. There'd be these price feeds constantly pushed on-chain so the state would live on-chain. And then, you know, your application could just read from like the latest price update. The problem is that's really expensive. And so it kind of worked with DeFi and like, you know, a few of the top currencies could have price feeds on-chain. But it doesn't...

Nicholas: So the currency would pay for the service rather than the like updating it. Nowadays, you pay to update it.

Peter Watts: Yeah. Well, I think it would be like the DeFi protocols were maybe paying Chainlink or Chainlink was just selling link tokens. I don't really know how it was all funded. But it just doesn't scale for NFTs. So our model, rather than constantly pushing data feeds on-chain, is more that when a user goes to do an action, so let's say they want to take out a loan against their NFT, they come to us, they get the latest price, they get a signed message from our Oracle, they bring that to the lending protocol, which can then verify that, oh, yes, this came from the reservoirs Oracle. It's a fresh data. So you're only updating the price feed on demand when a user wants to take an action. It's kind of piggybacking the user's transaction, which I think even Chainlink has switched to this model recently. And so, yeah, I think in terms of the use cases for it, yeah, I think. actually, I mean, yeah. So some of the stuff around lending has been one interesting use case. I think we haven't yet seen it, but you could do listings where I want to list the floor price. And so rather than me listing at a fixed price, and then having to update that as the floor moves maybe, or I could say I want to list it at 10% above the floor price. And so it's always sitting there at the floor, but a little bit above. if someone wants to hit it, then I know that I made a bit of a premium on the floor price. And you can do this maybe with a seaport zone. It's something we've looked into, but at the same time, it's like maybe a little bit complicated. And by the way, we don't just have a floor price Oracle. We have a floor price, we have top bid, which is maybe a better representation of the value of something. We have the flag status of whether something is disabled on OpenSea or not. So we have some AMMs. So there's always a problem on NFTX where people would dump the stolen tokens into NFTX. And so it was less interesting to buy from NFTX because you would be getting a stolen token. So some of the newer AMMs, I think caveat to this is they use our flag token Oracle to decide whether you're allowed to sell into the pool or not to avoid just getting stuck with all the stolen assets.

Nicholas: Got it. So in the function where you deposit NFTs, their protocol is going to check a signed message from reservoir to make sure that it's recent and indicates that it's not stolen. Cool. That's pretty cool. And I guess it's not extremely gas intensive to do that because it's just checking if a message is authentic. Exactly.

Peter Watts: Yeah, it doesn't add too much. I mean, it's a little centralized. You're kind of very, you know, you're relying on us. And so there's a bit of a liveness kind of issue. But this is our position on the Oracle. It's kind of like a fun thing. that's like, well, we've got the data. It's literally just, you know, people ask us for various data in our system and we'll sign to verify that. yes, this is the current state of it. I think that if sufficiently interesting experiences started to get built on it, then maybe we would invest further in it and figure out ways to make it a little bit more decentralized. And maybe it's kind of a kind of threshold signature amongst a group of people. Not sure exactly, but it's just like something we could put out there and let people experiment with and kind of unblock them and see what happens.

Nicholas: Very cool. I guess, you know, closing up, what are you looking forward to for the rest of the year and 2024? Is there anything in particular you're excited about shipping or changes you're excited about in the NFT market?

Peter Watts: Yeah, I think a big thing for us is definitely the multi-chain. And it's more than just like, let's add some more chains and you can choose between them. Like I think. finally, we're going to get to the point where wallets don't make you switch chains. There'll be a little bit more kind of just all that one. It's certainly like a challenge, but it's like a perfect challenge for us. We started abstracting marketplaces. We also now abstract currencies. We let you buy in any currency. I think abstracting chains is the next frontier. And so thinking about how to do that really well, thinking about smart contract wallets, the account abstraction and how that all comes in. There's all this talk around intents and how can maybe the actual people not pay gas themselves, but they're just saying, this is what I want. I don't know exactly how all that will play out. But it does feel like something we can get involved in and build some really nice infrastructure around. And ultimately, we're just excited about actual applications and users coming. So I think seeing some cool stuff around social, feel like it's starting to bubble a little bit. And yeah, hopefully, we can actually see some real more mainstream use cases now that we've got a lot of good infrastructure. now that block space is a lot cheaper. There's no more excuses. People have got to build that stuff. And so it's pretty fun for us to help them build that. So yeah, that's hopefully the next year.

Nicholas: Absolutely. Awesome. Peter, thanks so much for coming and talking about Reservoir. This was really interesting.

Peter Watts: No worries. Thanks for having me.

Nicholas: Thanks everybody for coming to listen. See you next week, same time, same place. Alright, see everybody.