 the Mellon security approach. Hello everyone, my name is Rita Trinkler, I'm the chairman and co-founder together at Mona Elisa of Mellonport and we are doing technology-regulated investment funds completely with ERC-20 tokens and obviously security is paramount for us. So I'd like to talk a bit about the Mellon approach to security and I hope you're all gonna see that we try to have a really holistic approach to security. So holistic means or like to give a short overview. I'm gonna talk shortly about what is Mellon, what are our values. Then a bit theoretical part of it, like how we designed the protocol. After that we go into code, so how we develop our code like from smart contracts to frontend, how we approach the auditing process and how we go in even beyond technology and try to shape regulation. So Mellonport, we're doing technology-regulated investment funds. So this concept of technology regulation is really this concept of contracts enforced by technology. So if you're doing as more contracts then you don't necessarily need to have a lawyer or an auditor or an enforcer. The technology is actually sufficient, the technology enforces itself. So we're using this concept of technology regulation and build investment funds on top of it. So investment funds is really just a set of assets, usually from a set of people, the investors, given to a manager and are holding in one place the custodian. The manager then can take those assets and trade them against other assets, preferably in a way that is beneficial to the investors. So what are our values? How do we want to design this? We want to design it in a way that is reliable. Reliable really means that it can't be undone. So since we're doing an alternative way of asset management, I think it's important to build it in a way that it can't be undone. So it's arguably what made Bitcoin to what it is today, the fact that it couldn't be undone. The second aspect of reliability is that it secure. So obviously if it isn't secure then it's not very useful. I think that's why we're all here today as well. So the second value is it should be permissionless. So permissionless really means that it's not for us to decide who can and cannot use Mellon. The same way it wasn't up for Tim Barners Lee to decide who can use and cannot use the internet. And the third one is it should be ownerless. So ownerless just means not controlled by a central party. Mellon should be a common good. The same at the internet is a common good. So Mellon. This is a short overview of the design of the protocol. In the middle you see a Mellon fund. A Mellon fund or like the core aspect of it is really just a set of smart contracts. This set of smart contracts acts as the fund custodian. So as I'm sure you all know smart contracts don't have private keys. So they can spend assets themselves. They need to have custom code on how to spend those assets again. It's really a core aspect of our security approach. The second thing those smart contracts act for is they are supposed to be or they are the fund accountant. Meaning they calculate the performance of the manager. They calculate the share price. They calculate the management fees, the performance fees like the incentive layer for the manager. And to do so they need to have data. So they need to have data of the prices of the assets. And that's where the data feed model comes in. So the data feed model is essentially just another set of smart contracts which deliver data to the blockchain. So we were very lucky with having Oracleize together with Crypto Compare building a data feed model for us that uses native proofs to validate data. Which means that data from its source to the blockchain is almost impossible to manipulate. Another great data feed module just recently announced is the one from Thomson Reuters. So you can have your Mellon fund now evaluated against data delivered from Thomson Reuters. Great. So we have fund custodian. We have fund accountant. If you look at what fund administration does, then we see fund administration does fund custodian, fund accountant, KVC AML and risk management. So we've just seen that we can solve two of them, fund custodian and fund accountant with smart contracts. So how do we tackle KVC AML and risk management? So that's where the participation module comes in and the risk management module. So for those that don't know like fund administration, they handle a trillion dollar, it's a trillion dollar industry. So okay, so participation. So the way participation works is really if someone wants to invest in a Mellon fund on smart contract level it's like a phone call to another set of smart contracts with a set of parameters and essentially asks the participation module is this person allowed to invest or not. So for example, we're working with the government of Zug that essentially then you know they work with U-port to have like to register citizens. So this participation module will be then like calls on smart contract level to the government of Zug and ask is this person registered with us? If yes, this person is allowed to invest. If no, then not. So but this really extends to just government of Zug. Like it's essentially every jurisdiction or every certain need can be implemented in such a module. And the fourth aspect of fund administration is risk management. So risk management is similar. So if the manager wants to make a trade it's like calling the risk management module before the trade happens. Is this trade allowed? Yes or no? So for example, one one example for risk management is only trades are allowed that are above the current market price. So that trading that is not beneficial for investor becomes really hard and really leads to a system where the manager and the investor they don't actually have to know each other. The same way if you make a bid transaction you don't actually need to know who's behind the public key it's sufficient to just trust in the technology. Does technology regulate the funds? Okay, so what happens if this set of smart contracts now needs to be upgraded? So essentially we have this concept of version which is like protocol version similar to Ethereum. All of the melon funds are identical in each version and if a version now needs to be upgraded then it's essentially just a governance layer that will be most likely just a set of people that are kind of voted in by the token holders so it's a small set of people thus they can act efficiently and fast if needed yet they're still legitimized by by all of the token holders. So yeah, those set of people then can add or remove versions effectively upgrading. Great, so now a bit about how we actually code these things. So we currently use TAP. So for those that don't know TAP suite is really a great tool. It's really nice for solid unit tests and they have like a husk implementation of the EVM. So it allows developers to really dig deep into the lowest level of the EVM to see where the buck is for example. A second great tool that we use is Oyente. So we're currently funding the development of Oyente. It's essentially a smart contract analyzing tool. It allows to test your contract against certain properties. So for example, re-entrancy can be checked automatically with this tool. Another really really great tool is you can build in custom assertions. So for example, you know assert that the accounting checks out after each function call then you can apply Oyente and try to violate that assertion. So if that violation isn't as if that assertion isn't violated that's a good sign that's actually not possible to ever violate that assertion. And we integrated that now with Travis CI so every time we make a commit Oyente gets run. Then we use just my tests. Just my tests is really useful for us for this holistic approach. So we just mind we can build like expectation sets and then share them across GitHub repositories. So for example, we have the same expectation sets on smart contract level as we have in MelonJS our JavaScript library. And we also just recently switched to ParityJS. So we have the ParityDevChain instead of TestRPC. So how do we approach frontend? So just maybe a short explanation. Like if you build smart contracts and if like 99% of your users end up using a frontend to those contracts then arguably the frontend is similar as important as or the security of the frontend is similar as important as the smart contracts itself. So that's why we use IPFS. IPFS is really great for security because it has this content-addressed nature. So content address really means that the content that you view on IPFS is addressed by its hash. So you can verify on our GitHub that in our build folder that the hash that we create is actually the hash that you access on IPFS. So there's no way to build in any malicious code, you know, code that for example tracks you or tries to steal your private keys or collect some data. So it's really transparent and it's really reliable as well because it's hosted on essentially a BitTorrent swarm. So yeah, makes it really reliable. Obviously permissionless and nonetheless as well. Then we have MellonMail. So MellonMail is essentially a messaging service that is as secure as your crypto. By the way, if you want to see the frontend, it's Mellon.fund. If you want to see that the mail service, it's Mellon.email. It uses your private key to encrypt your communication and stores it on IPFS. So the reason why this is important is imagine you work for Jamie Diamond and, you know, he threatens you to fire you if you ever trade with crypto. But with MellonMail you still can use, you still can have a Mellon.fund and, you know, having essentially just your public key exposed, your investors can still communicate with you and you never share anything more than your public key. So how we approach audits and bug bounties. So we start with audits from well-known and well- respected individuals and companies and then open it up to a general bug bounty so that everyone can participate. And eventually we deploy a live Mellon fund that is funded and whoever can hack it can have the amount that it was funded with. By the way, we always incentivize people. So if you found a vulnerability in our courts, no matter if the bug bounty started already, you will get incentivized for it. So as technologists we love what we do and technology is really important for us. But I think a holistic approach to security goes beyond just technology. So that's why we created MAMA, it's the Multi-Chain Asset Managers Association, which is essentially a trade body that tries to get innovative companies together at a table and allows them to give a voice so that the conversation isn't just dominated by the big banks or the traditional asset managers but that innovative people and companies actually can organize and have their voice as well in order to shape legislation and regulation. Great. So lastly, check out Mellon.fund, it's completely an IPFS, has a beautiful interface to it. Check out Mellon.email, go to orienta.mellon.fund, check out github.com slash Mellon project. We love pull requests, we love issues. Follow us on Twitter at Mellonport and join MAMA if you're here representing an institution. And maybe just lastly, the reason why it's good to go to Mellon.fund and start practicing it a bit is that we just recently announced that we're gonna issue 500,000 Mellon. I think it's worth like 30, 35 million right now, just in the context of fund management competitions. So you can participate those and if you're good at it, you get a lot of Mellons. If you're not that good, you'll get still a little but not that much. So that's essentially how we're gonna issue those remaining Mellons. And by the way, also on github, like we have, if you're an algo trader as developers, we already have like an open source trading bot that you can use and modify and you know why trade manually if you can build a trading bot. Great. Yeah, that's it. Thanks very much. We have time for one or two questions. Any questions? You have a question? Hello, first of all, congratulations. About MAMA, you're just working on Switzerland legislation or working with other countries like US or Singapore? So we would like to make this available for everyone but obviously our time is very limited and we can't make this compliant to every single jurisdiction. So we kind of started with government of two just to have like I guess a signaling effect that this is actually possible, that the governments and jurisdictions actually want this. But actually, like with the participation module, anyone can build those and we intend to incentivize like the people that build good participation modules and good risk management modules, we'll get incentivized for it. Yeah, I was so concentrated on my own question that probably this was a similar question but the requirements to KYC, whose requirement is it? Is it the fund manager who is only allowed to serve KYC customers or is it you as a software developer? No, it's the investors essentially. Like it's the requirements that are like what is required for an investor to invest. Okay, so if you want to attract regular investors, they need to prove to their, I mean the people who watch over them that they let the money manage by identifying managers. Is that the requirement? Because I thought the KYC requirement was for the investors, so the investors need to be KYC and it's not their requirement, so it's not their need that there is KYC. Because I mean I can be an investor. I mean there are investors who want to remain anonymous, right? Right, but you know that in a similar way that there might, there are probably a lot of managers that don't want to have anonymous investors. Okay, yeah, yeah, sure. Hi, Reto. Hi. The initial roadmap was the product to be delivered in two years, but it was early one year. Why was that? Yeah, so we're well ahead of schedule, so we're aiming to have this life with a basic form of governance next February 2018, so that's one year ahead. Yeah, but what happened that you could make it so fast, faster than you planned? So we really, like I think in all our communications, we're really not hyping, and it's just in the same concept, like we didn't want to hype and say, okay, we have this in two weeks, but we wanted to be very conservative. If you're faster and it's great for everyone, if you're a bit slower, then we still have some cushions. Thanks. Okay, one last question. Hi, I'm wondering how mature is the integration with the exchanges so far, and what's your roadmap on integrating with them? Yeah, that's a great question. So for now, we are using the exchange of WASIS tax, which is like a fully decentralized exchange, but we're working hard on integrating other exchanges as well, but also this, like it's all a matter of time resources, and we want to have like a working product, a life working product as soon as possible, and from then on, continue like integrating other exchanges.