Exploring Polkadot with Parity Technologies ft. Joe Petrowski

GM, GM. Welcome to what's going to be another fantastic episode of DevNTell. So if you didn't know, DevNTell is a 30-minute podcast held every week allowing founders, hackers, and anyone in between the opportunity to come on the show and showcase their product. And today, I'm ecstatic to welcome Joe Petrowski, who is the Director of Runtime Engineering at Parity Technologies. So if you didn't know, Parity Technologies focuses on the core engineering of Polkadot, overseeing all on-chain functionality and integrations such as staking, governance, bridges, smart contract execution environments, and cross-chain communication. So if you stick around for today's episode, you'll get to meet Joe, learn about Parity Technologies, Polkadot, and more. All right, let's do it.

GM, GM. Welcome to the show, Joe. I'm ecstatic to have you on today, man.

Yeah, thank you for having me. So I've listened to some back episodes and I really liked it.

Awesome, man. Yeah, love to hear that, love to hear that. And thank you so much for checking out the show. Uh, yeah, uh, so we are going to have an action-packed or sorry, jam-packed episode today. Um, I know that Polkadot has been making a lot of waves and they've got a lot of exciting things coming, likely very likely powered by the work that you and your team do. Um, but before we get into all that great stuff, would you just like to give a brief introduction about yourself if folks aren't familiar?

Yeah, my name is Joe Petrowski and like you said, I'm the Director of Runtime Engineering at Parity Technologies. And my night job is I'm also a professional skimo athlete. So I'm doing that a lot, especially on the weekends. And my career started in aerospace engineering, but I've actually been full-time in Web3 for I think nine years now. And if you want to follow me, you can follow me on all platforms @joepetrowski and I also have a website which is actually just my name; it's petrow.ski.

Awesome, man. And yes, um, for those watching on X or YouTube, Joe's X handle is on the description and the of the livestream as well, so definitely give him a follow. Uh, but wow, that that seems like quite a jump, um, going from aerospace engineering to blockchain engineering. Um, just curious, like um, what kind of fueled that jump for you and um, did aerospace engineering prep you for for blockchain engineering?

Yeah, actually I would it did in some ways, like definitely um, it's what where I got the introduction to programming and computer science. Uh, but then I actually took a little bit of a detour and I was a professional cyclist for a few years and I would say I learned a lot a lot more practical things as a cyclist than as an engineer, which uh, has helped me in in my career in Web3, especially just dealing with chaos and uncertainty and and bootstrapping things. Uh, my background in engineering was in like a a really large company and of course they've kind of faced every problem before, so there's a lot of structure and departments and stuff. And what you find in in a new industry like Web3 and in a startup like Parity is a lot of times you run into a problem and there's nobody really there to to help you; you're kind of on your own and you have to figure out how to do it. And a lot of my experience in professional cycling, like going to a new country where you don't speak the language and you know, something happens, uh, like you realize that everything is closed on Sunday and it's not just a little bit closed, it's really closed, and trying to get like food after a race or something, you you start to to figure out the kind of logistics and planning and and problem solving to solve these problems on the on the fly, and I find that those skills actually help a lot in the industry that we're in.

Yeah, yeah, definitely just because a lot of the things that are being built are net-new and nobody's really done it before, so you basically are left on your own uh to figure it out, um which is exciting, but at the same time also scary. Um but yeah, that's very that's very interesting to hear how um sport, cycling in particular kind of helped you uh hone those skills uh so to speak. Um but that's amazing, man. Uh and um yeah, uh you were telling me before we um got on the air uh that you joined um Parity Technologies um quite early on um about seven years, um I believe you said, um coming up in a couple weeks' time.

Just curious, like um what what attracted you to to Parity Technologies, um and uh yeah, uh just want to hear the story there.

Yeah, a lot of it was the opportunity to work on core tech in the Web3 space and especially as I learned more about the Web2 application stack and a lot of the trust assumptions that are inherent in it. I actually came into crypto via trading uh because my background in aerospace was in shock and vibration simulation and so this is very similar, it's like time series analysis and trying to make sense of seemingly random data and coming up with strategies around like kind of what is it going to do next. Uh, but kind of a few years prior to that, the Snowden revelations had a really big impact on me. I was working at a defense contractor at the time so there was this really a lot of really inspiring uh or invigorating conversations going on in the office and especially between generations like the younger people tending to find it uh actually oh no, that this was the right thing to do, whereas as you can imagine like in a US defense contractor the older generation was a little bit more like that guy is a traitor and uh I think these conversations really uh influenced my thinking and I I did join from a practical perspective more into the trading side and then when I started to to dive deeper into like well what is this thing that I'm trading and start to learn about the tech stack and realized how much it aligned with my philosophy and beliefs in this area. I wanted to to work on that and I got much more interested in building on the application stack than than just trading. And then I was looking for jobs in I think it was 2018, it was like a year after I had retired from cycling and I was trying to figure out what to do next and I was mostly applying for jobs that were kind of like investment analyst and writing research reports and it was kind of aligned with with my background in like applied mathematics and time series analysis. And then this position for uh I think it was a research analyst at Parity came up and it was like, you know, building, you know, greenfield technology and working on this this tech stack and it was definitely out of my comfort zone and I never really worked on like production-grade software before because doing like analysis is kind of like you you're not really running like for optimization because you can kind of run the job and come back tomorrow and see the output. So working in this type of software engineering was was very different but I was just really inspired by the mission and really drawn to working more in this vertical than in finance. And so I kind of rolled the dice and applied for it and uh ultimately got the job and now yeah seven years later I'm still here.

That's amazing, man. Um and yeah, uh you definitely made the right choice it seems. Uh, yeah, so um and over the years you you worked your way up to Director of Runtime Engineering. Um and uh kudos for you for for climbing up the ladder there. Uh just curious, like um what's your day-to-day like and what does the what does the team kind of concentrate on um working on the core protocol engineering there?

Yeah, at this point my day is a lot less engineering than it used to be. Uh, so I I had a phase where I got to do a lot of engineering and then became a team lead of a small team and was kind of dividing, you know, 50 to 70 percent engineering and then also managing the team a bit on the I don't want to say on the side; it was part of my job. But now I'm managing uh three different teams; it's a it's a department. And so it's gotten a little bit out of hand and you also have to realize a bit uh that you become a bus factor because I've tried a few times in the last one or two years to take on some engineering problem because I still am an engineer and I enjoy doing that. But then you realize that like you write the PR and the feature and then it still requires a lot of ownership down the road and you realize that you don't have the time to commit to doing that and it's actually much better for the organization to like to mentor somebody else and help them grow into that problem and take it on than to do it yourself. Uh, so kind of answering your question with a negative; my day-to-day is not much coding and engineering anymore. Uh, I work a lot on I guess like hiring and kind of like onboarding, running this uh department and then also uh we recently had a CEO change, so uh Gavin Wood, the founder, came back and uh of course has a lot of like some strategy changes and a future vision for where he wants to take the company. So a lot of working on our roadmap and how do we actually convert that vision into something that we can execute on. Uh, and that comes with a lot of work and cat herding uh as you can imagine. So I I would say that's my day-to-day job right now is is cat herding and yeah coming up with the the plan about how we actually do this.

Yeah, and uh yeah, that's it's no it's no easy feat, that's for sure. And and I feel you on that um, like as you get more uh senior-ed and tenured in in whatever role you're in um, you find yourself having less and less time to actually um do the quote-unquote fun stuff in in coding and you have to do more management and planning and strategy, um which which is fun in itself. Um but yeah, sometimes you get that itch, you get that itch to come back and and like cook up a PR or some feature you wanted. Yeah.

Coding gives you some like instant gratification, you know, when when you build the feature, when when the thing compiles, when the tests pass. Um when you're a manager, the the gratification tends to come in like 6 to 12 month cycles when when you've worked on a huge project and onboarded a bunch of people and then you see it finally go live into production. But you don't get so much of the day-to-day uh like joy of creating something new.

Yeah, 100%, 100%. But um you still push, you still push for nonetheless. Um and you mentioned it kind of there um with uh Dr. Gavin Wood. Um just curious, um have you gotten a chance to work with um Gavin a lot? Um maybe in the early years of you joining Parity or um maybe now? Um just curious, like um how is it like working with someone like him?

Yeah, it's a lot of fun. Uh, I worked with him a lot especially in my first few years here. So kind of leading up to the launch of Polkadot uh we were working together a lot. And then I believe he took a break of about a year or something for for some personal things. And then uh for the last few years he's been working on this a couple projects actually like one of them is uh Proof of Personhood, which he's talked about at a few events like the Web3 Summit, and this other one called JAM that he's also talked about a lot with a lecture series. And the last few years we were kind of working separately; I was working more on like the existing Polkadot tech stack and he was working on like the vision for the next iteration of it. Um, but now yeah in the last two or three months we're working together a lot and of course like any leader uh and you know he pushes you a lot and there's some pressure with that, but it's a lot of fun because there are always, you know, a hundred new ideas and uh sometimes it's a bit of work to prioritize and get down to like well these are, you know, the five that we can deliver in the next six months. Um, but yeah it's a lot of fun and always kind of thinking about the cutting edge and and the vision and what comes next.

Oh I bet. Uh, yeah he's he's one of those um special folks um in the space. Um and yeah that must be really exciting to to get to work with him like day-to-day. So amazing, man. Uh yeah, uh yeah. Um so um shifting gears slightly, um I want to get into Polkadot um specifically double-click into that. Um so um in your own words, um what would you say is the elevator pitch for Polkadot for people who might be watching the show today or or whenever and maybe new to crypto and Web3?

Yeah, of course it depends on your background and how you approach these things, but our goal with Polkadot is to build a protocol that can host fully Web3 applications at scale. So as an example, like something we saw earlier this week was the Cloudflare outage and a lot of applications that purport to be Web3 went down, right, because of this centralized provider um or centralized component uh going out. Uh so it kind of shows that a lot of things that are Web3 are still kind of reliant on trusted service providers and our vision for Polkadot is to build a tech stack where you can build a a fully trustless uh application and that means storage, sybil resistance, no reliance on centralized providers. Um and there there's a lot of work and hard problems that come out underneath that when you really dig into trying to do it in practice. Uh but yeah that's the 30-second pitch.

Yeah, and um I mentioned at the start of the show that and you kind of hinted at it as well with some of the things Gavin's been working on. But there's a lot of exciting things kind of happening on Polkadot. Um, especially uh since the first time I I took a peek at it a little while back. Um what are some what are some of the more exciting things um in your mind that are uh kind of helping boost the ecosystem today and maybe some things that you you and your team are working on?

Yeah, so we're kind of at an interesting point or an exciting inflection point where the last few years we focused a lot on solving the quote-unquote hard problems. So we've had like a decentralized sequencer from day one, right? Well we call them collators but in principle it's the same thing as like an L2 uh sequencer. Uh we have on-chain governance that self-enacts and it's not just a signal. We have a really robust staking and staking platform and security guarantees that apply to all of our parachains, again like that being similar to to an ETH L2. And we have focused on those problems largely and consciously to the detriment of builder experience, developer experience, user experience, and that's why I think you don't see such a vibrant ecosystem on Polkadot compared to what you see on like Solana and Ethereum L2s where you do see a lot of hype but they they are also running into the quote-unquote hard problems like decentralized sequencers or having a Cloudflare outage to take down MetaMask, right? And so um I think we're finally actually at a point where we're really happy with the lower level of the stack like the governance, the staking, the validation, the scalability of parachains, uh some of our execution environments. We're really happy with those and so now we are starting to turn our attention toward product development and making that experience of building on on Polkadot better. And so we're starting now with this like prototyping phase where we're trying to build products ourselves within Parity and that's something very new for Parity where we have mostly lower-level protocol developers. And so we don't really have something flashy to to show for it today but the whole point of this exercise is to really expose what are the pieces that we're missing and I hope that in, you know, 6 to 12 months we can come out with this full stack where we say here's how you can make an application, we have decentralized storage, we have sybil resistance, you can run light clients on your phone or in the browser, in an extension, you don't have any centralized RPC host or provider, there's no centralized storage, everything is happening in a cryptographically secure way. Uh there are no central intermediaries that you are depending on and uh I don't have a concrete answer for for something for what those components are now but we're in the process of figuring them out and and I think it's a really exciting direction.

Wow, yeah, indeed those sound exciting and um yeah, um that's a good tease, I think that's a good tease for our developer audience to get excited here. Um and uh still kind of sticking on the um the developer experience here, um like just curious, what is the smart contract development experience like on Polkadot? Does it differ from um the the EVM side of things or the the Solana smart contract development?

Uh, with EVM the answer is yes and no. So we do have our own RISC-V based virtual machine called PVM. And about 18 months ago we started this project to integrate this into Polkadot. And the idea was to have uh the PVM uh virtual machine but still make an Ethereum compatible tech stack. So we have uh a compiler called resolvc which compiles Solidity code to the to the RISC-V PVM code. And then we also worked on all the tools like Foundry and Hardhat and uh and Anvil and like everything else that you would expect as an Ethereum developer. It would just be compiling your bytecode to PVM instead of EVM. And what we learned along the way was that this worked for new contracts and the newer versions of Solidity, but there were a lot of legacy contracts like Circle CCTP comes to mind as like an obvious one that are on older versions of Solidity and they just do not play nice with RISC-V at all. And the the thought of kind of asking all these people who have contracts that they can just kind of deploy to rewrite their contracts and have them re-audited, it's a lot of hoops to jump through. So uh we also added an EVM back-end, so we actually have one interface that has two different virtual machines as a a back-end. So if you want the full 100% EVM Ethereum experience, you can use the the EVM and just deploy with that as your virtual machine and everything is basically guaranteed to work. And then uh but we have found that for like more compute-heavy operations including for ZK applications, the PVM performs much better. And so if you're writing an application from scratch and you're not really beholden to like an old version of Solidity or something and you have a bit more freedom, um then you can still use Solidity, like the newer versions, because we have this resolvc which will compile it to uh PVM bytecode. But you can also write contracts in like Rust or C or something else and you know if you really want to like optimize the performance for a specific compute-heavy application you can do that. So the tooling around PVM is is still pretty new and I think we'll have a lot of room to grow there, but that's also something where in 6 to 12 months I'm really excited that we're going to have something that will be able to have a really high performance uh smart contract platform and that will enable a lot of new on-chain use cases that might not have been possible in in EVM environments.

That's that sounds super exciting, man. And yeah, uh can't wait, uh 6 to 12 months out. Uh, you think it's a lot of time but it's not, it's not quite. Yeah, it goes fast. Yeah, it goes fast like that. Yeah.

I think it will be live uh we were hoping it would be live before Christmas but it looks like maybe in the first one or two weeks of January this will all go live. So uh I have a few links like in drop uh into the show notes and that has documentation on our kind of trial version on Testnet but if you're really if you're really keen to to start going without the the PVM specific tools then we have some like you know docs for nerds that we can share and hopefully some people can can get started using it.

I'm sure, our our audience is geared towards cutting-edge and playing around with with different tooling. So we'll definitely have that documentation there for folks to mess around with. The weekend is coming up as well, so no better time. Uh, and um you touched on it a little earlier earlier on, um uh but I want to double click a little bit into a JAM um that is still it seems like it's still in kind of like a research um development phase. But just want to hear like um what is JAM and and if you can provide kind of a high-level overview of of what it is for the audience today?

Yeah, uh so JAM stands for JAM or Join Accumulate Machine. And uh I don't want to disappoint because I know your audience is very developer-heavy and I haven't been super in the weeds on JAM. But I I can give a high-level overview of it. And the main thing is that when you look at if you look at like the original idea of first I'll start with like from ETH to Polkadot. If you look at the original idea for ETH2 it was having a bunch of shards of the Ethereum chain but still with the same runtime environment, the EVM, right? And then the fundamental thing that we changed in Polkadot was to actually make them heterogeneous. So we said, you know, instead of forcing the EVM as the you know quote-unquote back-end uh of each chain, we'll just as long as you can compile this to WebAssembly you can use it. And the idea was that people could write their application uh in a more custom way including smart contract environments, but also if they have a specific use case and they don't need smart contracts, then they can just put all that as a first-class citizen into their runtime. And what we found was that this was also a little bit of a one-size-fits-all model. It's actually a parachain is actually overkill for a lot of applications. A lot of applications are just fine as a smart contract. Um they don't need the the horsepower of a full chain. Uh so it was a bit more like infrastructure and demand on the application developer and if you didn't already have product-market fit and didn't know what you wanted to build, it was rather burdensome to go with like the parachain route. Um and we also found that this like asynchronous composability was a big burden for people because a blockchain is like a it is like a single-threaded computer, right? So you have a bunch of single-threaded computers and they they kind of have a way to talk to each other, but they have to wait at least one block. So they have to finish their execution, you know, hand off their their state root and they can send off a message as part of that and they're but they're going to have to wait for a response which could come a few blocks later. Um and that's a little bit different than how like a CPU, like a multi-threaded CPU, would run an application on your computer. So the idea with JAM is to make it feel almost more like a like a normal like if you're using like a 64-core MacBook then make it feel a little bit more like that. So um it just gives a lot more flexibility in what the services that you're running are. And so I can give like a few examples like one is like you could run smart contracts directly on JAM rather than um having to have like a parachain and then like using a smart contract on that. Um and you can also do the scheduling a lot better. So rather than having this oh it's going to run like every six seconds and you need to like get a spot in this block, you can just kind of like buy this allocated execution time. Um and then finally it it does allow for somewhat synchronous execution, so you could have two applications that are um in principle separate, they don't share memory, but you could get them to agree that like we actually want to be scheduled to execute together and they can make synchronous calls between each other in the time that they've been allowed to execute. And then they can go once they've done their their synchronous calls they can go back to being their standalone applications and doing independent scheduling. So um and then it also has quite a bit more scalability, so uh I don't know the details of the scalability plan but there there is some plan that we can kind of clone multiple JAMs and share security across them uh by sharing a validator set. Um so yeah not the best answer, I know, but it should kind of point people in the conceptual direction of it.

Of course, yes. And yeah, again, like sounds like there is going to be a lot of exciting things kind of coming up in the Polkadot ecosystem around JAM and some of the things that you mentioned earlier on. So um like how can how can people um keep up with with all this? What's the best way to follow along?

Yeah, I think following our work uh you know on our website, on GitHub, on Twitter, you know we're sharing there whether it's from Parity or from the I think it's called @polkadot on Twitter. Uh so yeah I think we'll be posting all the major updates there.

Amazing, amazing. And yeah, those details will also be in the description of um this podcast um so definitely um you'll want to follow follow along there for all the latest updates. And um just as we're about to close here and since we are a um developer audience um mostly, um are there any opportunities that um Parity Technologies is hiring for that you might want to highlight on the show today to catch someone's attention?

Yeah, Parity is hiring. So we we have opportunities I think for quite a few roles. So yeah I work in the runtime, which is we're also hiring because we have a few improvements and projects that we want to make at the runtime level. But also as I said earlier, a lot of our work is focused on prototyping products and then building out the infrastructure to make the next generation of products easier. So um especially if you're interested in like um tooling, testing infrastructure, uh product development, API development, anything that makes building applications easier. Uh storage is a big thing, so like working on storage. Um we have yeah a lot of gaps to fill and so yeah especially on this like integrations and tooling uh and API design fronts there there's a lot of work and then also product development. So uh I think we're looking for a few in-house product developers and uh maybe some UX people, so there there's a lot that especially in like the upper part of the stack.

Wow, man. Yeah. Uh and and all those opportunities can be found on the careers page of Parity Technologies?

Yeah, that should be kept up to date.

Excellent. Awesome, awesome. Well there you there you go, folks. Uh what better time uh than now uh to dive into the ecosystem of Polkadot and start playing around. And uh why not, why not throw your hat in uh to see if you can join the join the party yourself. Um but uh as we're coming to time here, Joe, just want to thank you again so much for taking the time out of your day uh to come chat with us. Yeah, thank you for having me. It's been a lot of fun and uh yeah I look forward to to seeing it. Yeah, yeah, my pleasure, my pleasure. And yeah, with that I just want to wish everybody a very happy Friday, very happy weekend wherever you may be, uh and safe travels for any of our viewers today who might be traveling back home from uh Devconnect. Uh and we will catch you back here next week for another great episode of DevNTell. So then, have a good one, folks. Cheers.