Jason Goldberg I Building a Startup in Web3 | Airstack I Pragma India 2023

[Music] for our next interview I want to bring on Jason uh Jason is the founder of airstack uh if you're at the hackathon you're going to definitely hear about it if you have been a part of the hack that you've probably already heard about it but uh we're not we're not going to talk about what a stck is we're going to talk about what it was like to start a company in crypto especially coming in from a lot of startup experience from doing things outside of crypto we're going to have an amazing conversation around all the things that goes on into building a company so let's bring on Jason on stage to talk about starting a startup in web 3 hi everyone so um Jason we've known each other since before you started your stack and um when we met you were at a company and you started a company called OST so I want to start off by actually um having you tell all of us about what that company was and what you wanted to do and and when did all that happen yeah so first of all I've been an entrepreneur for 20 years um and the first company I founded was a company called social media back in 2008 and uh the idea of social media and was enabling uh basically the news feed before we had Facebook Newsfeed and Twitter timelines and um so I've been always been enamored in so with social and in Networks um I had a unicorn e-commerce company called fab.com which actually had a big Development Center in Puna um so I spent a lot of time in India um over the years and in 2015 I read the ethereum white paper and I was like this is it this is like the next phase of the internet we can program uh you know smart contracts we can do programable money we can start to build the financial internet um start to reshape applications uh and so I became obsessed with this and so I decided to start a company in 2016 uh which the intent of the company was to build the first decentralized social network um and we 201 2016 and we called it poo pepo and we we first built a centralized kind of messaging Community app and then we said okay now we're going to start to decentralize this um before we even launched it to users like we just like got people's feedback on it they liked it and then we just realized that wow we were way too early um and so we pivot the entire business into something called OST which was open Simple token um and we spent 2016 2017 2018 2019 uh writing protocols for building layer 2 application platforms uh we wrote a smart contract wallet uh way before anyone knew what account abstraction was um we actually uh were the first to write something that you could Bridge tokens from a layer one to a layer two and then use the layer two and then reclaim back on layer one they're Merkel proofs um but way before anyone was ready to use this stuff yeah that's wild yeah so so what happened after 2019 yeah well um I mean one is co hit right and what happened with Co hit was we actually we're finally first of all I got to back up like I'm like a total like ship it and iterate kind of guy like my entire e ethos is like get something out there as quick as possible get people to touch it and give you feedback and it was frankly kind of depressing for me being in in web 3 in 2016 to 2019 because we were building stuff that no one was using I was completely you know against everything I stood for and we you know we we actually relaunched poo we launched it as like a like a version of it at uh I think it was e waterl in 2019 we launched at Devcon in Osaka we we showed it off won on a ux award but no one cared it was just way too soon um but so when Co hit we actually were just on the cusp of like actually going to be doing some onboarding we actually had a a deal with a company called Brew dog which is a a brew pup with like thousands of locations around the world and we were going to roll out a loyalty program for them um based on our our technology we were working with Microsoft on a bunch of things with inperson events we were working with grab to launch like there actually grab has actually done some of this now but we had actually worked on the first kind of iteration of a grab uh in Singapore kind of wallet for like loyalty points when you um riding cars and then Co hidden everything we were working on was in person um and it was just like it's not going to happen total shut down not gonna happen and so actually what happened was one of our investors said hey you guys are awesome at Building Product would you build a virtual Fitness platform so he did that for a year and a half um and then sold that and then jumped right back into web 3 um after that yeah that's uh that's an incredible last six years um what was kind of the obvious lesson from uh starting a lot of the social stuff back in 2017 to kind of 19 you kind of said it's early but what kind of gave you that impression that hey this may not be the right time yeah I I'd say like first of all there's there's for myself at least I think for a lot of entrepreneurers there's no worse feeling than being right but being too early um like we so much of the stuff that we wrote and that we developed back then has actually made itself into other products today um there's bits and pieces of open St in polygon in near seen it in like the in the the cell white paper you dig in there's actually pieces of our U like the way that we Define that you would set up like a layer two and first the side chain um and uh so that's satisfying in some ways but it's also sad that we didn't get to do it right um and I'd say like you know one of the things for me is like I bowed to I never want to build that way again I always want to be building with a customer in mind and building product that people want to use and want to use right now um and that's really something that I've decided firmly is something something I'm always going to be doing from now on the the startup world has an analogy around like say it again so the startup world has an analogy around kind of vitamins versus painkillers and uh and kind of it seems like being early sort of feels like a vitamin 2 but you don't know if it's actually that or you're just early and and maybe would have worked exactly the way you thought two years later yeah um what then kind of led to air stack yeah so so my co-founders and I we were seeing co-founders as open St um um we we knew we wanted to build something in web 3 we we really you know we're just fundamental Believers in the power of these Technologies and we got back together uh thinking about it the end of 2021 and we said okay if we were to go back and solve one of the problems that we identified back when we were building op St um and validate that first that it's actually a real problem worth solving if we could find something like that that we can then go spend you know the next 10 years of Our Lives on that's what we wanted to do and one of the things so we you know battered around ideas for a few months and but we kind of kept going through like what were the pain points what were the problems that we saw developers going through and the one thing we kept coming back to is what we saw was that uh at least back in like 2019 and 2020 we saw that when people were starting to build daps obviously the technology was all very early but what we saw was that everyone was basically building the same infrastructure over and over again especially if they needed to connect their daps to any data information on chain um where it integrate offchain data whether it's from ipfs or elsewhere everyone was building the same exact infrastructure and going through all the hassles and headaches and pain to do so and so we decided to see if that was still a problem so we went out and interviewed hundreds of developers uh end of 2021 early 2022 and we really said all right what are your biggest frustrations what are the things that are stopping you from going faster from building better applications and what we kept hearing back from them was was all these well same thing of like well I'm you know I have to spend so much time indexing building a database to analyze the data um I have to figure out how to get you know data from from here from here from here and then we also heard was that as more and more protocols daps and blockchains become usable that the problem for developer just becomes harder and harder and harder that it's one thing for developer to say okay I'm just going to build on one protocol and I'm just going to learn the In-N-Outs of let's say polygon and I'll keep up with their re and I'll keep up with everything I need to on that chain but once you start going to the next one and the next one and the next one it was like almost impossible for them to keep up so then we had like this spark like the moment of okay okay if we could be this infrastructure layer for all of these application developers who want to be able to build apps faster easier um make them more composable by integrating data from across blockchains across daps across protocols you know we could help them go from you know a Year's worth of headache and heartaches and AWS bills to you know snap your fingers integrate it and keep moving then we would be both a painkiller and a vitamin and frankly some speed as well and that was that was the goal that we set out to do yeah that's that's an interesting analogy it's a tweet quote right there um well so maybe maybe I'll ask this more directly for everybody to have a better understanding too what is airstack now yeah so um so we built that right so what we did is we spent uh 2022 building uh this infrastructure layer API platform for developers to very easily integrate uh onchain and offchain data into their applications to do so cross dap cross blockchain uh at literally snapping your fingers um we our mission you know we set out is just to be developers best friend and really understand what developers need to help them build better apps and uh move faster we launched a private beta in January of this year um we had about 180 developers in the private beta we ran for January February March we got their feedback on it iterated iterated during that time we also built an AI engine um because we realized that our uh apis that we had built that enabled this kind of cross cross kind of querying um were uniquely suited for training in llm and so we actually went from a developer having to let's say spend you know a couple days or a week learning airstack apis previously 6 months to learn how to index all this stuff we went to literally within milliseconds with our AI you can query and say for instance show me everyone at pragma who also attended a previous pragma their ens name their forecaster profile their bio get their photo from lens and etc etc you can do all that in our AI in a matter of milliseconds and plug it right into your app with a Json file that's the output and um it's changing lives frankly I mean it's like developers are able to build in you know in a day or two they would have to build you takeen them you weeks and months before I mean that's that's super powerful recommendation just generally kind of coming out of a box um private beta was q1 you kind of went out there publicly in in April yeah what's going to happen over the last uh nine months yeah um so we launched at eth Global Tokyo um that was very purposeful uh we my own experience back in the space in 2018 2019 I found the eth global hackathons and events to be the best in the business for serious Builders and so we set out a strategy of okay we're going to ship fast iterate ship fast iterate you know kind of go through this process and we actually signed up to do every eath Global hackathon this year this is my seventh um and also myself and but what that forced us to do though so we did Tokyo in April got feedback then we did uh Lisbon in May got feedback iterated we did uh waterl in June got feedback iterated we did Paris in July we did another hackathon New York in September and then we're here and so it caused us to have these like really rapid development cycles of like you know put the product in front of developers get their feedback do it again and again again and around July time frame we just saw like magic in the bottle hit right that were you able to trace what what was that moment yeah uh what it was was um we built something um so we have all this integrating all this data from across blockchains and we built this abstraction layer on top of it that we call the onchain graph which enables you to literally see here's the people that cartek has the most in common across all of his onchain behaviors for the last six years um and to do so not just with a single wallet but across multiple wallets that you might have used over the years and kind of associate them and what happened was I mean true story Shane at xmt saw it and he's like this is the future this is what every xtp app needs to have um as the address book the contacts in every X SMTP app uh started introducing us to everyone who's building xtp then uh stonnie from L did the same thing Dan from farcaster did the same thing and yeah the net output is that in the last eight months we've done over a billion API calls amazing a billion a billion and 700 million of them were last month okay yeah we call that product Market fit yeah yeah and it's I mean it's to be honest it's it's humbling but I feel like we've earned it because we've worked our asses off to do it yeah well congratulations this is this is incredible bu Stone um and that's just that's just hug getting started and really grows exponentially from here yeah and the thing is when you do something like that like now we have so we have right now 490 customers right which is crazy right and they ask us for so much and our whole ethos is around listening to them and finding out what devs need and so now the hardest thing for us is actually prioritizing what we do next because like they like the the worst thing a startup can do is lose focus yeah so one other interesting thing I just realized from from your last answer is that it seems like a very common theme Here is is recommendation and kind of seeing social behaviors being represented in in quable you did that with OST you did that with pepo and now with the the recommendation here for airstack um seems like web three social was still top of mind for you yeah um so maybe my my question will be now that you kind of focus more on the developer Centric product um how do you define web3 social yeah so I actually so yes I'm I'm very obsessed with social but I have a different view on web3 social than I think most people do um most people when they think about web3 social they think about kind of like within the web 2 construct of like there's a social network and the difference with web 3 maybe that parts of it are decentralized so they think about the foraster network or the lens Network or maybe this cyber connect or something else and they kind of look at it as like okay it's a it's a social network that's built on you know the web 3 kind of rails right and my view is completely different my view is ethereum is the social network and we're entering a world where people don't participate in one platform that ethereum and all the ebms permeate when you use your wallet and you earn let's say poops from attending events there's over 7 million poops that have been collected now over six years of data of the events that people have been to that's a social network that's a social fabric that you can start to connect People based on where they've been and who they've been been with with who they've been with um when you join a DA or whether you participate in nft communities that's a social network now the fact that you also have let's say use certain daps there's people who also use those daps so we can kind of look at that and build social networks and inferences based on that as well and so I think we're entering a phase and this is like the onchain graph that we built is really kind of feeding into this we're entering a phase where social is a an identity and kind of recomend recommendations is something that's going to be permeating across all of ethereum and so it's not going to be that you just like yeah you might join a social spefic a social networking platform where you have like a friends list and a follows list but I think the exciting thing is that next layer on top of it of like everywhere you go with ethereum you can have recommendations and personalization if you want to based on your onchain behaviors just served up to you from the start and this is such a big departure from web 2 in web 2 there's no way you could build an application like developers doing with airstack there's no way you could build an application for instance said I'm going to take the social graph of Instagram of someone else's data yeah yeah and someone what else is doing on Tik Tok and like what Spotify listing behaviors and their you know flip card purchases and put it all on the single app you can't do it they won't let you but in web 3 you can it's just been really hard until now um and what that enables now is you have these whole new kind of genre of applications where they can leverage all of this to create something where there's no cold start there's a friend list from the beginning there's recommendations and things you can do right away and I think this is just going explode app development and I think it's going to explode user experiences where we this going to be a catalyst for seeing more and more people building better richer apps and then more and PE more people entering the ecosystem I'll say one last thing on this is people are always like well when it's easier to use more people will join you know web 3 and I say no when there's actually reasons for people to join they will hop through millions of Hoops and then but things will get easier along the way and so when we build better applications that give better experiences than what people can do web to all the rest will follow yeah I mean that's kind of the bill something people want wantt try right um I do want to come back to actually starting startups here um as you kind of thought about airstack from kind of all those learnings from from OST what are some lessons that you are now applying when kind of redoing everything from airstack versus that that you thought were potential challenges during OST yeah I mean I would just say it just comes back to making sure you're building something that people want something that they need and people that they're willing to pay for um and I think all too often in web 3 because there's a lot of like we're building Frontier Technologies in many cases that it's easy to lose sight of that and I think if you lose sight of that too long it's very easy to get it wrong and we got to get out of this cycle in my opinion of having lots and lots of web through applications who build really interesting science projects but then go under after a couple years because they can't find product Market fit and say get back to you know really focusing on how do we build products that people really really need and really are willing to pay for that's and that's what we do every day at airc yeah like I tell our team like our job is to talk to customers every day understand what's their frustrations are and to build them you know painkillers vitamins and speed I'm a I'm big farman now this is big farm IED um maybe on the controversial side what about uh having a token um I'm not saying no but I'm also not focused on it um and and but I will say maybe what some lessons learned kind of managing that for an OST so at OST we had a token um we were one of the very very few token projects back in the day so we launched our token in 2017 where we had a functioning version of the protocol first before we launched the token um that's rare for the people in audience um and there was a a use case for the token um in the protocol now the problem was just that being a use case wasn't enough because it was be years before there was actually a product that needed the token from a kind of end user standpoint um and I'd say so what am I like I people ask like what do I think about tokens today and I actually think that the SEC in the US like the Howe test is actually really right um and I think it's really right so the how we test is that there has to be utility before you sell something otherwise just a security right like a token um and I think that's right because what happens is if you sell a token and the token is going into the hands of investors and speculators um and the actual real reason for the token to be used has not been clearly defined and mechanics have it worked out yet then the token is going to fail it's very hard to overcome overcome that now it might succeed from a price standpoint as speculators just bid it up for a while but of actually serving the purpose that it's supposed to have inside of an application is not going to work yeah maybe let's kind of go back on thinking about building a bigger business so before OST you were doing faap yes um tell us about what Fab is yeah what were you able to accomplish and and what happened yeah so for those who don't know fab.com was a e-commerce Marketplace um we launched it in 20 2011 um it went rocket ship overnight uh we were at one point one of the fastest growing companies on Earth we were unicorn within 18 months did 150 million in sales in the first 12 months of the company um and things were going great and then about two two and a half years in Amazon just crushed us like literally crushed us um and when I M like Amazon crushed us at the beginning of Fab uh we would find independent designers and feature them on our Marketplace and it was typically about 45 days before they would appear on Amazon and by the third year of the business within 48 Hours of us finding a designer they were on Amazon with free shipping free returns in the warehouse and everything and we just couldn't keep up with it and it was like pain it was painful because we we had both the exhilarating ride of becoming a unicorn super fast and building a product that people loved and working with designers and helping change their lives to then really struggling to become be be relevant after that yeah and that's always tough um and uh it's even harder when you were able to accomplish that a decade ago not just kind of because everything is part of a bigger economy now recently how do you think about competition in web 3 yeah yeah um I say like you even with that experience with Fab like I love building products and what motivates me every day is our team trying to build something that people love to use and um you know making a dent out there with you know something that helps move the industry forward we really do believe in decentralization and the possibilities of this technology um I forgot the question I'm sorry what's kind of a good way to think about competition here like is that are we at the same skill where like we have to be Cutthroat or are we still come by out you will have competition at a certain point you've got to do better than your competition or you won't exist um so yeah like we're I'd say we are fiercely customer Centric and also fiercely you know competitive in terms of we want to be the best in the space and like by founding team and I we our vision here is to be a company for 20 years that's having a major impact in this space and so we have to constantly think about you know what could possibly derail that no that's amazing um let's kind of switch gears to building products but for developers yeah sorry last couple of uh last couple companies you've done have been more on the consumer and social friendly kind of individual retail side now Engineers are your customers and uh working with Engineers is a sometimes a binary situation yeah um what kind of have you noticed building products for for developers yeah I mean I would say I'm a product manager at heart I am deeply technical in the product I've been building you know products for so long and so it's it's passion of mine and I really want to see developers build better products in this space in particular uh I would say that going you know being in this industry in the previous bare market and seeing just the amount of energy and dedication that you know developers put in to try to building things that are even many times too soon like really inspired me um and I saw the real need to help developers build better faster in this space uh and so said okay if we can provide some of our you know consumer product mentality towards that that that could be a good thing as well and so our team when we build things we really try to pay attention on um both building a good develop developer product but also developer experience so like a devx as well as building no code versions for people who maybe don't know how to code and canou our products um so that's something uh and then we want we wanted really to build something in this space that we would get immediate feedback on and I think developers is like where to be when it comes to that now that said so if you I don't want to do a commercial here but if you go to airstack we have this thing called air stack Explorer which is a no code version of our tools that's built on the apis and it gets 5 to 10,000 visits every day for some reason people like it so it's a product it's pretty good yeah are there things that um you think will be beneficial for other Founders who are building de tools uh that they can kind of impart on on them yeah I would just say like make sure you're building something that people actually need um I am obsessed with customer interviews um and the one question that I constantly ask people is what frustrates you the most um because if I can figure out what you're struggling with then I can try to build you something to help you get past that um and if you're building something as a developer in the space try as much as possible to avoid building things just because you have a hunch or that you think it might move the needle validate validate validate and if you ever want help thinking about how to do customer interviews give me a holler happy to help amazing um well let's let's try to kind of wrap this up so there's a lot of stuff that you've kind of been through um and a lot of it is good this is uh this is not meant to be like this just an observation and um and I'm finally glad to hear that airock is kind of meeting the success and people are actually trying it and you're actually becoming a a critical layer in that infrastructure of of enabling a lot more web three applications some of this has already been kind of uh kind of said for me already so aside from that vamin painkiller Speed Line any other parting advice to Founders who are thinking about building in the the we 3 ecosystem now in the current market yeah I would say that like you know you're lucky right now to be building web 3 because we're about to see so much of this stuff become usable and I would say the biggest takeaway that I have from the eth global hackathons this year is you people are Dow able developers are Dow able to build complete applications on a weekend and it used to be that you'd build like a little PC that still needed like you know years of work to bring the market you would shout at your computer for the first day and then building Begins the next day yeah and so like for instance like like now at the when when air stack we're judging projects we tell everyone no videos it has to be live and if the product doesn't work end to end we're not going to consider it and because it's possible now and so now we're on the cusp of this next w wave of I really think we're going to see in 2024 and 2025 kind of an explosion of actual usable uh onchain applications that are going to not just be copies of web 2 but can be entirely New Primitives and experiences that take advantage of the various layers of web 3 um from you know decentralization permissionless to uh composability to you know trustless all that kind of stuff and brings it all together in applications um that change people's lives and I'm excited for that awesome well thank you so much for that amazing conversation and uh hopefully some of this advice makes a lot more new unicorns next year thank you please give him AIG Round of Applause thank good