Rethink Trust 2018: Workshop 4, Provectus Blockchain Framework -- Trustless Cooperation
Recording: Rethink Trust 2018: Workshop 4, Provectus Blockchain Framework -- Trustless Cooperation
okay well if you're here you definitely want to know about how to properly build a blockchain a blockchain based distributed application generally speaking and this is a good place to be because evangelists would come here to you know in the right setting but and that would be typically forward-looking aspirational etc we are here to be a little less aspirational and forward-looking and no more retrospective and rearward looking to discuss our experience in actually things that have worked well things that haven't worked well and what did we end up building as a set of capability of framework to help businesses to help companies construct distributed solutions that are realistic that are functional that will work in the real world production and scalable environments so it's not aspirational but hopefully it's practical I was talking downstairs to somebody and say imagine somebody standing up here and talking about how tcp/ip will change the world or SQL will heal disease and yet that's what happens with blockchain today people talk about blockchain as some kind of a magical wand that you can wave and somehow it's going to do all of this magical things yet blockchain distributed Ledger's much closer to tcp/ip and SQL and such utility systems and tools than anything else and that when when we finally get to a point where people start talking about blockchain in the terms of that they do as tools such as tcp/ip and such as SQL we know we have reached the point of real ballistic use of this technology and proper and appropriate functional use of this technology for enterprise purposes well I'd like to cover today and I will turn it over to our team members who built the framework is we will ask Andre on to actually cover the components and experience of building a framework and why we believe generally speaking those components are a minimum must if you're going to build distributed applications and then we will ask Alex where is Alex they're different Alex but this alex is always willing to talk as well to touch on some of the use cases and we will take some Q&A if there are any left over at that point we're starting a little late so we'll try to get through it in time as scheduled so without further ado Andre on please it's your your floor yeah Thank You Yuri thank you everyone everyone for coming and what I'm going to talk today I'm going to talk about the work and the framework which which our company is is implementing but before talking to about the architecture of our framing I would like to to show you the picture how it started how this story started before a few years ago before starting to investigating and before starting as a real implementation of the framework we needed to understand what actually blockchain technology is we needed to understand advance advantages and disadvantages of the technology the main advantages are quite obvious it is decentralization and it is no censorship it is trust less environment but we needed to understand what what we should pay for this advantages and actual gold chain technology has a lot of a lot of challenges which you need to do to deal with actually can anybody say the disadvantages of the technology of the blockchain technology a lot and say okay do we know more it's slow okay something else not exactly scalability perfect very good yeah there are a lot of myths the way a lot of myths about technology I've heard a lot of time that such a phrase is led that will put everything onto blockchain we will put our data into good chain it'll put everything in it because we'll chain it is immutable and our data will never be lost we will calculate everything with blockchain the blockchain will check everything a smart contract will go there he'll go there you'll check and when do the and does the work no it is not it appeared it is not true about the technology and actually the first the main disadvantage that we came across technology is that it is expensive I did a small investigation a few a few days ago above the current prices to store one kilobyte of the kilobyte of data into a very Impala clock chain and it appeared to be $7 for one kilobyte the same story is for Bitcoin yes if you use if you use private permission and blockchain you can decrease this costs because there is no cryptocurrency but you will still pay with your computational resources because the data is replicated across all the network and the same process the same calculation is done for each and every transactions it that that is there so you'll still pay for it and if you try if you'll try to put a huge data on blockchain you will you will see that it is likely to be impossible so you mean you need to to solve this kind of tasks in another way for example you can use traditional database store your data there and just you just store a hash of this data and to blockchain and it is enough it is enough to reach immutability in this data if that if the big size data is properly replicated in a decentralized way the next the next challenge is that transaction commit time is slow for public blockchain like it theorem it is 15 seconds 4 bit Bitcoin it it is even a lot more as 10 minutes and you can reach you can decrease this this time for when you use permission and watching but still you will not you will not be able to reach speed for each transaction less than several seconds it is due to restrictions of the network protocols you need to reach consensus for each and every transaction that means that you need to have deferred work for each transaction that you do on the invisible chain for example we had a use case when we tried to transfer a small small pieces of data via blockchain for one specific use case and our test was our test environment was generated boh-chan blocks each second but there was a communication between peers which which needed to transfer their data several times if I am not mistaken it was four times in each direction and the whole process was completed in about 1/2 min so you need specific you cannot reach solve these tasks with blue chain technology only you need something like to deal with a queue of tasks you need you need to use other technologies like message brokers for this decision-making process the network is quite slow because dela there are some times of decisions that can be can be done only in democratic way only in decentralized way for example if you need to change the version of your smart contracts to change the version of your core logic you cannot do it by the via the central administrator you need to agreed get each and every each and every member of the network you need to get agreement and as far as it is agreement it is usually it is voting then this this takes this process takes takes a lot of time this should be done with with appropriate processes incentivization of the chain network should be well thought out because this network has no single point of control that is why we have a tragedy and tragedy of Commons here because it is a shared resource for example one of the one of the members of the network can consume a lot of resources as it is depicted here the member can consume one of the members can consume 10 transactions per second some of the member does not consume at all this can happen for example if some some of the members would like to have tea those type of attack just over just fluid in the blockchain with a lot of fake transaction and the blockchain network will stuck in that case in public boot chain it is it used it is usually sold via a small transaction fee for each and every each and every transaction but if you do it with permission at blockchain you need to solve it on your own on your own because because you don't have native native cryptocurrency and in general what chain technology is is quite new and we have no we have no enough experience in the world of using blockchain in enterprise area so we don't know which which type of blockchain is better it should be used better for each case and that is why we are trying to build our framework to be pluggable and to use any type of low chain so this is the main principle that if you figure out no chain should be used can be used for decentralization but know no more than that because of all of these restrictions this is the architecture of single node so what we address what can be built with our framework we can build a network of of companies network for companies and each of the companies will have this stack of of technologies in a single node as you can see except the blockchain pier itself you can see you can see traditional database you can see message broker which is rabbitmq you can see cryptographic key management it can be it can be useful if you need to have compliant cryptographic storage it is just an example how can it be because the low level low low layer is pluggable you can you can add additional additional types of blockchain for example the low layer if you can use permission and block chains as well as public machines you can add for example you can use permission with chain to store to store some specific metadata which you shouldn't need to share with with the outer vault which you need to share only with your partners and you can use you can additionally use separate separate module at the low layer which will be connected to the public book chain and you can use it to payment for example via cryptocurrency we have we have high level API is here and with this high level API layer it is possible for developers and for companies to implement business logic and do it in a fast and efficient way because all the complexity of all the underline and align layers is is hidden and we already have several integrations with real real partners and we implement in business logic uses specific layer so an example of communication of two companies can look like this the companies have count blockchain network it can be public book chain it can be permission about chain and the data the data transfer is can be done via direct peer-to-peer communication between companies nodes so in that way we can avoid all the restrictions that are there with blockchain we do not put we do not transfer data via blockchain we do not use it to store huge data because the data itself can be stored in a traditional database which is MongoDB for example and as soon as this data should be can be shared or should be shared to another company you can you can write a specific logic to shared but in addition to using complementary technologies to blockchain we were thinking about the support for block chain management because it is very it is it is really really a big problem to manage block chain networks you should think about a lot of way from the ropes and the wops point of view and we have specific we have support of this process we have automatic develops DevOps tools the automatic scripts and then boarding process the process when you need to add additional partner or remove additional partner from the network is done almost automatically you don't need to know anything about in the line technology you don't need even to know anything you don't need even to be an expert or developer it is seamless I think this is it related to architecture we have the current current stage of our framework is is a beta version we are ready for for real integrations but we have plans for the fusion the primary the primary option for blockchain is the darién based network currently we use proof of authority type of blockchain but we can easily migrate to public as their own blockchain if you have plans to in our roadmap to implement other types of machine maybe hyper ledge may be exonym or naam blockchain frameworks we need to complete performance in stress stress testing with different types of blockchain we need to implement VPN it is a business process and notation notation - we are going to implement it on top of business logic and with the tool that is possible to use graphical to graphical tools - to work with and to create this business logic come versioning system of smart contracts data management service which is a central service should be completed as well it is on our roadmap and we have we have plans to to investigate and to implement decentralized storage modules in our stack I think decentralization it will add additional additional possibilities in terms of decentralization yeah this is it do you have any questions yeah please Alex can you share with us some product management experience yes okay okay we did not because we do not have proof of work as a primary option we I think this this problem will not appear in most of the cases first of all what our first our first option is to use proof of authority aetherium consensus protocol so we there is no such type there is the such type of a text still is possible if more than 51% of members will agree to to fake the network but if more than 51 will agree you cannot do anything here but it is not the case it is not the case for many networks and then the number the numbers that use that you mentioned there's another case that can be the case we are building a framework if you need for your use case to to have more than one blockchain network in the stake it can be done so for example if you need to to transfer data from aetherium to hyper ledger for example that can be done if that is users keys if you remember the picture there was a layer of api's it was the second from the bottom yeah we need we need to put some an instigation for specific type of blockchain we need to implement API and the higher-level if you remember will be unified a program a program yeah maybe but there are there are specific type of operation for example put a hash on the blockchain vote for some decision that can be done I'm sure that can be done with any type of the chain it is already done it is possible to to access the lower if you need a more granular control you can access a lower a lower low level of api's and do what you need good question we have a basic logic when first of all each of the members which is unbothered has is written into specific smart contract so no one from these specific members can make any transactions and as far as if anyone would like to play not by the rules it is as far as machine is transparent you can see it right away instantly and if you remember there is of boarding process according to our custom our default logic this misbehaving party can be of boarded by collaborative decision so it is quite easy to implement actually not only API layer but whole core logic is is done with no GS not just technology in non GS language we chose this language because it's quite widespread technology widespread language and it appeared to be good good choice because before that we we had intention to use other technology I will not say which one thanks Sandra and guys let's give it up for Adrienne for his awesome talk and actually for the tremendous jobs that Adrienne and the team has made in order to make this thing happen and actually in order to empower us in preventers in the product management division to be able to dream big and actually to work closely with our partners and customers on addressing their issues with help of decentralisation and blockchain and today like in nearest ten minutes I'd like to walk you through three pretty different cases where our framework and the technologies can be used and they have one thing in common framework for sure and one more thing which is the digital Walt which is the personal digital world which is owned either by the client the individual or the organization or by the community so by some entity so for instance let me start from one case which is about the things we hear in news which is about journalists and their work is military zones and non democratic societies so we know that there is a problem of journalists working in so such zones trying to do their job that they intend to but they're being found because they're not anonymous and they being put in jail killed etc and there is ik and we our partners we're searching for the solution how to address this issue how to address issue of allowing them to post and to get money for this and we have come up with this solution of providing those journalists with their personal digital world where they can post and put their articles and with the tools which are made into the world using the modular structure of the framework they can not only store it and not only can unanimously send it to newspapers local newspapers local media and international ones but get paid for that because mamie's are where options you know to anonymously send the tip about something but you could never get the credit for this you could never get some money for this so it was never working and democratic journalism was not really possible in so zones so digital world personal digital world with security with functionality inside can handle it and we hope that will be released and will be changing the landscape of the countries and of the news we know about yeah then let's give it up for them because that's a very important thing one more thing which we are thinking it's about the user putting user in the current business models and inca system in the middle not the enterprise like now but the user and getting control of the data back to the user so what we're thinking we we're dreaming that just an individual can have it's his own digital world which has his all data his personal data his data about usage of his driving records about you see his medical records all kind of e-commerce records and having all this records under control we help the user to have them and monetize them and we help the enterprise's to comply with all these government regulations so this is a basically win-win solution and keeping in mind that our architecture is modular and those modules can easily interact with the world we can provide those users with ability to for example to send their data unanimously let's say to insurance companies they say there are medical records they're saying their finance records they're sending their driving records and they get the best possible quotes for insurance because insurance companies are just know everything about this person and can actually decrease the risks and may and and helps the data profit their individual so the data becomes a really digital asset and the third thing is actually about not individual but organizations organizations may have their digital waltz as well and let me show you the example let's say what we have a customer is this a real case we have a marketplace we're on the first side their tax consultants which have SMBs and big enterprises to work out their text reports and we all know that together text report you as a company you should give very sensitive information where your revenues coming on what are your plans or your owners all the things which you don't want to share with anybody actually and that's why some companies they hire big departments of people who are working like one months in a year but getting paid twelve months in a year just like not a perfect mathematics for the enterprise that's why we have come up with idea what if we say them you keep your data in your digital world and those text consultants just work with the data in your digital world and they cannot get it to themselves and the awesome things that they actually this candy here that not only that text consultant can get this information but the marketplace they don't have access there because it's all decentralisation and it's control over the data for the organization and our individual in the core so again the concept of digital world keeping the data having some modules which help you to communicate with the world it's in the core of all these cases which we can empower with decentralized technologies with blockchain and we can accomplish them quicker and a more efficient way if used with our pro victors blockchain framework thank you any questions guys and one more thing if you can you have some ideas of the business models like this you are really we appreciate if you just text us give us a word because we think that blockchain problem is in tech maybe and we're trying to fix it but it's even more in how it can be used not in 10 years or 100 years when we apparently all debt but now in the current enterprises in the companies which are funded now or will be founded next months