Mapping the New Frontier of Personalization: Decentralized Universal Interest Graph -Ethan Champagne

foreign [Music] thank you [Music] foreign [Music] thank you [Music] all right good morning East Denver and uh thank you for coming to our presentation my name is Ethan and I'm the business development lead in Partnerships with first batch where we are building the universal interest graph for every app and so a little bit about our mission most as you know most web users are also social users and on web 2 platforms we're able to get recommendations for Content people to follow experiences based on our social graphs and these social graphs don't yet exist in a unified way in web 3 and so what we're doing is we're making this personalization through the universal interest graph self-sovereign interoperable and privacy preserving and so just a little bit of background about engagement with apps that do have personalization versus apps that don't so Tick Tock is arguably the best recommendation engine in web 2 and they've reached a billion users faster than any social app and this um this personalization is what enables Tick Tock to create a better ux for their users and Escape some of this avoidable user churn that other platforms experiences without personalization and so this avoidable user churn just to put a figure to it is costing U.S businesses alone about 136 billion dollars a year and this churn can be reduced by up to 67 percent if companies can succeed or even platforms can succeed and providing a positive pers first experience with their users when they come onto the platform and so additionally in web 3 users are coming into platforms with Anonymous identities suede Anonymous identities and there's not the same sort of depth of social data that's available in web 2 in web 3 to personalize a user's experience um and that leads to users spending more time on platforms lower conversion rates and overall you know deteriorating the ux for the user which is something that not just web 3 socials but all web 3 platforms need to address and do better and so our protocol um starts with the layer one which is Danny which is a decentralized approximate nearest neighbor search engine and so Danny exists in web 2 socials and content platforms that you may already know so Facebook's recommendation algorithm fights Spotify is recommendation algorithm annoy these models we've taken and made them available open source on top of our weave via Smart contracts and so any protocol that's interacting with first batch can query Danny to provide personalization to their users or users can query it directly with their zkid the second layer of the protocol is our indexing API and so this happens in two ways the first is the indexing of user Persona data so first batch is indexing off chain public apis and can transform interest interest traits about users into vectors that can then be uploaded onto our Hollow DB decentralized identity database on our weave and can be queried again via Danny and we are also vectorizing content on web3 and web 2 sources to be able to make them available for matching with IDs and so the third layer of the protocol is Plug and Play curated feeds so any through any platform it doesn't have to be web 3 socials specifically but any Marketplace nft you know any kind of shopping experience we want to make Discovery and personalization available to any app on blockchain or off chain as well and so where again the elements of our protocol are open source on top of are we so we provide a ready back-end for platforms to integrate first batch onto their front end for personalization with about two lines of code foreign and so I've mentioned before a zkid and a large part of what we're building is about user privacy and so in order to get the same personalization in web 2 you usually have to forfeit your data rights to the Platforms in order to have your experience personalized by their social graph but what if one day your favorite social platform you know is no more you know your your social graph disappears with their centralized servers and so with first batch ID users connect their social accounts via a ZK off which basically gives our AI permission to go and provide and conduct an analysis of users public off-chain data via the apis so Twitter YouTube Reddit Discord and we're able to vectorize these Persona traits and upload them onto our weave on the Perma web so your zkid will live you know regardless of what happens to the platforms that you use you can always have your ID on our weave and the great thing about ZK is that we're able to abstract your persona from your real ID so we don't ever sell user data we don't work with marketers we don't do ads the focus of our project is really to provide personalization on these platforms and so you can users can feel secure that we will never expose their data their data is safe and so with Hollow DB again our decentralized database that we've made for ID query management built on top of our weave we've reached web 2 like speeds for querying so there's users don't feel that much of a difference when getting the same service and we've managed to do that through zks which are pretty lightweight and the last part is self-serve and personal self Sovereign personalization so as I mentioned before users and platforms can query Danny to get their personalization on any platform and so this is just some snapshots of our ux as I mentioned before when users log into first batch ID they will see a screen like this that will prompt them to connect socials that we like Reddit Twitter Youtube Discord GitHub the next step is first batch AI generates novel Persona embeddings based on those connected socials when the process takes you know anywhere from 15 to 30 seconds but once that is complete we've scanned all of your off-chain socials and you'll be able to see your persona ID visualized in first batch ID platform and so you may see here that there are traits visible to you as the user but this will never be visible to anyone else who is trying to provide you personalization so platforms other users won't be able to see you know your specific interests and a unique thing about first batch is that you can choose users can choose which trade they want to be made available for hyper personalization so users can control who how much and when their data is shared yeah and the in the long term our goal uh is to be able to integrate this decentralized personalization engine into any platform users can sign in with first guys instantly to games and nft platforms e-commerce web3 socials live Commerce and eventually web 2 platforms as well to get hyper personalization instantly and so yeah first batch Persona ID is live now anyone can sign up connect to social to get your zkid and in the coming weeks you will be able to right now we also have decentralized content feeds that are available so with your first batch ID you can get recommendation for your favorite RSS views from web 2 decentralized social lens diesel farcaster mirror and you will also be able to log in with your first batch ID to your favorite decentralized platforms so thank you for your time and attention and yeah take care thank you any any questions from the crowd thank you all right thank you [Music]