# Aztec Labs | Private DeFi: what we can build today? - Lisa Akselrod | ETHDam III - 2025 - Speakers: [Lisa Akselrod](https://streameth.org/speakers/lisa-akselrod) - Channel: [CryptoCanal](https://streameth.org/cryptocanal) - Date: 2025-10-07 - Duration: 20:03 - Watch: https://streameth.org/watch/yt-wsbgfQbuyjc - YouTube: https://www.youtube.com/watch?v=wsbgfQbuyjc ## Description Welcome to the 3rd Edition of ETHDam, hosted May 9–11, 2025 in Amsterdam. This year, we brought together the brightest minds in privacy, security, and AI for a unique 48-hour hackathon + conference combo. 🌷 https://www.ethdam.com// 🌷 ------------------ Aztec Labs | Private DeFi: what we can build today? - Lisa Akselrod | ETHDam III - 2025 🎀 About the Speaker: Currently DevRel/Tech. writing in Aztec, previously in Taiko. BSc in pure math. Leading a math book club "zero knowledge considered harmful". Occasionally launched a memecoin $SINGLE. 𝕏 Follow: https://x.com/cryptobuilder_ https://aztec.network/ https://x.com/aztecnetwork ------------------ About ETHDam & CryptoCanal ETHDam is powered by CryptoCanal, an education and events platform rooted in Amsterdam, expanding into Rotterdam and ZΓΌrich. Keep up with us to see updates on future events: https://www.cryptocanal.org/ Follow CryptoCanal on X: https://twitter.com/CryptoCanal Join CryptoCanal TG Community: https://t.me/CryptoCanalCommunity Join CryptoCanal Discord: https://discord.com/invite/XJVjpCqQBz CryptoCanal unites crypto enthusiasts committed to making a positive impact. Unapologetically political, we prioritize education, events, and services while championing cypherpunk values like privacy, sovereignty, and censorship resistance. ------------------ πŸŽ₯ Credits: Intro / outro by babyPRO - https://babypro.art/ ETHDam Photography by Paulus – https://concretestate.eu/ MC of ETHDam - Laura Brown - Cofounder at JobStash & Veri, and your resident crypto ginger. https://linktr.ee/laurabxyz ------------------ Special thanks to our partners who made ETHDam possible: 🌹 Hackathon – Bouquet: Oasis Network https://oasisprotocol.org/ 🌷 Hackathon – Petal: Circles https://aboutcircles.com πŸ’› Conference – Gold: Zano https://zano.org/ Dash https://www.dash.org/ Bitvavo https://bitvavo.com/en 🩢 Conference – Silver: Igra Labs https://igralabs.com/hero πŸ’› Conference – Copper: Lido https://lido.fi/ DeTrip https://detrip.travel/ Cake Wallet https://cakewallet.com/ The Grid https://thegrid.id/ Calimero Network https://calimero.network/ 0xbow https://0xbow.io/ Mina https://minaprotocol.com/ JobStash https://jobstash.xyz/ Cyber Capital https://www.cyber.capital/ POAP https://poap.xyz/ Acronym Foundation (Supported our Top 10 Hackers) https://acronymfoundation.org/ 🌱 Sponsor: EF Ecosystem Support Program https://esp.ethereum.foundation ------------------ 0:00 | Intro: Private DeFi Vision 1:01 | Transparency vs Privacy Tradeoffs 2:26 | Why Global Users Need Privacy 3:17 | Tools for Building Private DeFi 4:09 | Private P2P Payments & Lending 5:21 | DEX Privacy: AMMs vs Order Books 6:38 | ZK-MPC for Matching Trades 7:32 | Private Stablecoins & Collateral 8:29 | Prediction Markets & Stealth Tools 10:45 | Use Cases Beyond DeFi: Voting & Social ## Transcript Welcome to East to E to E to E to E. Hi everyone, my name is Lisa Axelro. I am part of part part of def uh defi de team in adtec and today we are going to talk about private defi in general when we think about like blockchain good blockchain bad what should we build on blockchain and we have a lot of opinions usually like um some people think we should empower sovereign individual some people think we should empower pro-democratic governments some people think we should care about user data even though it's not really clear if the user cares about data. But one thing that we are more or less have consensus on is that like traditional finance is bad and like we can do better. The question is how better we okay thread five bad we will do defi But there is a nuance. This is a nuance. When you start building defi on blockchain, the blockchain is transparent. So say okay, we will fix traditional finance by defi that is permissionless and 24 to7, but you have to put all your data on chain for the whole world. minor nuance. So we say, okay, welcome into the fight. It's incredible. You just need to put all your data on chain. In general, the trade-off is not clear. Are we better off or worse off? I would say that worse off. What we do uh by that improve surveillance. Great. Add data to the analytics of corporate giants. Fantastic. Show everyone that you have no clue of making savings and spent again half of your salary on random Great. This is what I think about that. But the thing is we can do better. If you want wide defi adoption, defi should be private. Otherwise developers are quite constrained in what they can deliver because they can't give users guarantees of their data. Many jurisdictions just do not allow from the legal perspective to put u users data public. Many countries have dictatorships and surveillance governments and people just will not chip in into any transparent defi. So trai is bad. We can do better. Our better is private defi. Today we are going to talk about um what exactly with the tools that we have today what exactly we can build in private defi because in fact it's not it's not that transparent. So we'll assume that we have three things. Um we have universal zk language noir that is high level uh rustbased zk language and we have blockchain with both private and public state and this is very important because you will see it very soon when we think about what we can build it's usually the blend of private and public objects just making everything public or everything private does not work quite well in terms of the which products we can produce. And finally we will assume that we have a cosnark library. Um pretty much it means that multiple parties can produce zk proofs in a way that none of them has access to uh private input. So let's start the simplest thing we can imagine a crypton now banking style product like P2P transfers we get balances that are private we get identity that are private um it's settled in a way that it's only the receiver and the sender who are aware the transaction took place at all and the transaction can be made private up to the level someone did something on the network then we can extend it to other P2P services whenever the deal can be settled between two parties like for example for P2P lending lender and the borrower and it is possible to make it private in this case up to the level that two parties are not aware of other other parties identity so I'm a lender I'm u giving someone uh money I have no idea what is their identity but I'm still we have all the effect guarantees that the deal will be settled uh fairly and everything will be fine and the rest of the world is not aware that I'm taking a loan at all or like that I'm lending money then of course we have dexes with dexes it's uh interesting it depends a lot on the dex design first let's think about amm style like unis swap style when the user is trading against uh the liquidity pool um in this case the truth is we do not have too much uh flexibility in terms of privacy. The best thing we can do is making identities private because uh out of the design of the AMM um the assets the amount all has to be private because uh the asset prices are updated based on this information. But um even with what we have we can have an AMM that yes shows you what is happening but you have no idea who are executing all these deals. So it's already step forward. Thinking further we can do uh batch trades if someone is aware something like penumbra style uh where we process the whole block of transactions the whole um block of swaps. Um but here also there are nuances like what if it is just one transaction in the block then everyone can see what this transaction is. But again already some flexibility. Then we think about limit order book uh dexes. This is when we have uh buyers on one side, sellers on another side and uh someone trying to connect uh what uh both parties want and to settle the deals. This is interesting because in this case we can go up to something has happened. The transaction on the chain will look like something has happened and we will utilize the cosnark or what is called zkmpc network for it. Uh how it will happen? We will have this ZK MPC network who um can check if uh buyer seller pair uh coincides and can be settled but it will be checked in a way that none of the ZKBC network participants um has access to content of the deal. Then we have can have private decentralized lending against a bit uh again a bit tricky. It's close to like AMM style where we easily can do identities private and then again we need all the rest of information to be public to be able to make liquidations. What we can do as an alternative uh a bit more permissioned alternative again using cosnarks using the KPC network we can uh appoint a set of small set of validators for each particular loan and this set of validator will uh check the position for health. Then we have private stable coins. Uh it's like arguable if we put stable coins to defy or not to defy because most stable coins are not really decentralized and uh but just thinking about them in general again we can have private stable coins and here this property of something public something private um becomes really interesting and meaningful for us because on the user side we can say we can have it like identity private balance private it. But on the stable coin protocol side, we can have some transparency like total stable coin supply and proof of collateral. We want all of that to stay publicly verifiable and this same will uh be absolutely fair for asset tokenization mechanism. Um we can have uh it can be issued, distributed, managed privately and even the very existence of the asset onchain will be private. But we can leave something to be publicly verifiable if we want. Uh we also can have private derivatives in the uh same manner as tokenized issue tokenized assets. Erh oh the thing I really like is uh private prediction markets. Uh why I like it because it seems that in general prediction markets have already uh proven themselves to have some product market fit even with like full transparency and maybe introducing privacy it can go even further to like web two to style consumers. Uh but what we can do in terms of privacy we can have user identity bets and payouts private while the rules of the game and outcomes to be publicly verifiable. There are several mechanisms how we can do it. We can use something like stealth address or like one time address where every in every game the user will participate using a new address and the price will be distributed to this address. So if someone is uh looking from outside at the whole like what is what is happening within the prediction market they can't like do any analytics understanding like that those particular batch of addresses belong to one user. We also can use uh something called escro contract. This is self-executing contract uh that let's say can have its own balance and whoever can prove that they satisfy some rule for win to get the prize uh can get the prize. Yeah, there's other several DeFi products that from my perspective can meaningfully benefit from privacy and um really interesting to explore both from the like product perspective but also for people who like fancy text tech from the technical perspective. Um, one of the thing to mention is that like again it's a bit arguable if TVL is a good approximation for market size but let's say if like we already with transparent defy could get some 300 billion let's say market it sounds quite good especially like when people are quite depressed about blockchain at all And um it seems that like if we introduce privacy can't promise you but I think we can go go way better maybe 2x 10 x 100x I don't know. Yeah. And uh the last thing to say is that how I think about it like I often need people who say like we should go all in privacy like everything is terrible everything is transparent we should go all in privacy. uh how do they call it like cypher punk? Um I do not support this like extreme point. I think we should have privacy as a tool for building products for development for having fun and then whenever we think about like building something we should have this opportunity to use privacy as a tool. I think this was the last slide. Yeah, my contacts. Feel free to message me anytime to talk about random stuff. And uh yeah, this is uh the QR code for getting started with Noir because whatever you want to do with privacy, Noir is like your 100 bet. You will never regret if you open this link. That's it. We have two minutes for questions if we have any. Well, um, perfect. We've actually had a change in schedule a little bit. So, the next talk will start at 12:30, so we might have time for a few more questions. So, oh, question number one. Where? Oh, go here. Then here. Hello. Um, hi. I was last in the Aztec community call and they were talking about outsourcing uh proof generation and they were like you shouldn't do that but I just saw in your talk you were talking about coarks does that enable the user to outsource his proof generation because you can split the witness and not reveal your witness I guess to the rest of the network. Yes, I think um this is the most like proper way I would say to outsource uh proof generation because if you do it you still get your kind of like complete privacy definition where by complete I mean that it's only the user who has access to his private inputs because also like the definition of privacy is a bit tricky topic. Someone can tell you that like okay if it's only the user and the prover being aware of a private witness and the prover we are running it I promise you it won't tell anyone it's in general like web two model like where you have the server of the company and the company promises you not to share it with anyone uh someone will tell you this is also privacy so like privacy is a bit definition sensitive topic but if we are talking about like strict definition of privacy that we stick to in adtech. H as far as I know using coarks is the only the only way for today. Okay, thanks. Thank you. Question over here. Hello. Great presentation. Um I'm wondering if you could give a bit more color about for example onboarding someone like USDC. I assume you want to do that and I think you touched on this briefly but um I mean USDC they want to be able to like blacklist addresses and like kind of freeze funds. How exactly would that work on Aztec without going too much into the technical details? Yeah, that's a great question. Uh one of the features we have an ad tech is we do not have canonical bridge and uh it's up for every application uh how they what the rules of uh on ramp of ramp they want to bring. So if someone wants to have a a stable coin where users will be able to have privacy but there will be some criteria of who can use it based on KYC or whatever. Uh they can define the rules on their own. Uh I think it will be very close to how stable coins work today or you can even think about it maybe how to how um centralized exchanges work today like Binance where they have full control of like who is using it but uh with privacy for user uh it's uh like it's not it's not decentralized stable coin even though you also can build decentralized stable coin too like some fancy fancy experimental designs. But yeah, did I answer your question? We can talk about it after the talk. Perfect. Do we have any more followup? Perfect. Question time. Hey Lisa, so have you been thinking of any other use cases besides DeFi for public private state? for example, government, gaming, messaging, social media, like there's a bunch obviously. Yeah. Um I think whenever like you want to have some mechanism with uh many users versus entity and you want this mechanism to be fair like for example voting system set up by users not set up by entity. you can benefit from it because for example you do voting voting is private uh count is public publicly verifiable or you do crowdfunding donations are private um accountability of the contract is public thinking about it from like very pragmatic point of view in real world very often entities do not want it to be verified able and they benefit from it being like uh speculative. So to think about it not like theoretical projects but like real world projects um it needs this like very sensitive fit where the system for some reason has incentive to be publicly verifiable. Such a sad but true statement. It is sad statement because it removes a lot of use cases. But speaking about like uh for example social I think it's even more tricky because I think two or three years ago there was huge narrative where everyone was telling people want to own their data. It's just like how Twitter can remove my tweet and then okay a lot of people start building like all the forecaster stuff and then there are like 10 users okay maybe 100 users but incredibly small amount and every day you see someone tell telling oh I am moving from Twitter because it does this and this now you can find me on Farcaster and next week you still can find them on Twitter because everyone is on Twitter and it's very network effect dependent uh technologies is so seems that the problem in general is not as sharp and adding the privacy doesn't help too much and it's very also very interesting question because um within our long journey with like Ethereum from 2014 there were a number of use cases where like it seems it is good to bring blockchain then you bring blockchain it it doesn't really work out and then the question is can you fix it with privacy. In most cases, the answer will be no. It's just not a fit from the product perspective. But in some cases, the answer will be yes. So the most interesting thing that wasn't possible like a year ago and is possible now to find these cases where privacy really fixes the problem. Cool. Do we have any more questions or No. Okay. If if you think of a question after the talk then you'll be roaming around for the rest of the day I assume. So yeah, thank you so much. Um thank you so much. So the next talk