Save Corroborator Validated with airSlate SignNow

Add signed corroborator

recording son well thank you all for showing up today and I don't know drew if you wanted to come on and do a quick introduction and welcome if you say hi to everybody yes is per Miller and this is the blockchain applicated we just started had a couple free things and we focus on business applications aneema buys that really glad to have an expert like dan come present tonight I'm glad to be here I wouldn't necessarily attest to be an expert but I'm I'm definitely been here for a while learning it's be fun sounds cool well with that then I'm gonna go ahead and share my screen and pull that up all right everybody should be seen what I'm seeing and I'm gonna go ahead and pull up by slides so corroborator is what we'll be talking about a little bit tonight so uh as Drew pointed out thank you very much I'm Dan shields and I will also have Ian Phillips on the line who will give little demo and hello later and together we are corroborating on this project corroborator and really I'm excited to tell you about this is a practical business oriented use case for blockchain technologies so really corroborator revolves around this problem statement is how does someone trust that mission-critical data is uncensored and unalterable Incivek data is absolutely final I'm absolutely secure all right integrity meaning that we can validate it hasn't been manipulated with or tampered with and provenance meaning we know all all of the the W's the who what where why when how all of that has locks into this data and so if we want to have those things for others to be able to check corroborator x' can come to our date right so we've been doing a lot of market research and we've been studying this problem and it seems that there's a Common Market needs that come kind of surface when we're talking to people is for lust reliable access to secure file systems and data fingerprinting logs so understanding where the data is in this secure file system and who's accessing it is a huge issue right it's very common systemic throughout a whole bunch of different industries and from the people we were talking about primarily for responders as well as legal affiliated use cases that we'll talk about a little bit later on one of the biggest things that they would potentially benefit a lot from as an offline first architecture meaning you don't have to talk to some server or some cloud service somewhere that if you have a local copy or instance that you can continue to work and synchronize on when you get back online that's a much better way to go about things we're also very interested in exploring we're done in backups and this can be with trusted peers and providers where there's a whole bunch of especially with secure file systems that are cloud-based access to that data can become lost or worse can effectively become a blackmail situation where provider says well we're gonna restrict access to your you know files until you pay your licensing dues or whatever it might be and this becomes especially a problem for things like police departments and legal people we're also really interested throughout what everyone we've talked to in fingerprint and timestamps that are now logged and absolutely incorruptible but more importantly have a basically a hundred percent uptime so if these kind of systems go down or these access to be able to validate these fingerprints of the data are inaccessible and we'll explain some examples a little bit later on you lose a huge amount of opportunity here and basically like work shuts down on some of these systems go down for people so having access to that that's basically undeniable and uncensored bol understand uptime is really important and secure data sharing across this file system and access controls are really really important and then especially when we're talking about transitioning out of an organization's when you cross organizational lines that data control needs to be really really monitored internal it can be a little bit lighter although we feel it probably shouldn't be so architecting for the worst case scenario where everyone's an adversary is kind of what we want to build on and then across all of this data sharing read receipts are really important so this is what corroborator is really after is open data verification is kind of the subtext to all behind what we are after and I think the right thing to think about what we're building here is effectively with a middleware right it's pluggable in the sense that you can bring your own whatever system with engagement might be so this might be mobile applications this might be a IOT devices this could be websites wherever it is and corroborator wants to help plug you into systems of record that are highly reliable highly secure so things like blockchain and secured encrypted file systems and things like this we want to be able to bridge that gap so that as a developer especially it's extremely easy to take your app and plug it into this architecture you don't need an assist and to super well so it's customizable based on your use case based on your needs is what we're building so blockchain in file sharing and fingerprinting systems can all be effectively isolated once you can specify what your needs are and we can help you design how to use corroborator to get that output and expose a very simple API for you so in as I was alluding to you before designing to fit the most challenging use cases first and foremost where everyone potentially is an adversary is what we really want to design for at its core and then loosen up based on that dependent to get more efficiency out of the system right so uncorrelated at the individual records level is something what we really want to be able to do such that these fingerprints can't be associated with other fingerprints that a particular user is making making it totally private anonymous to begin with and then we can use any database right so especially we're talking about block chains right is the key thing we want to have these fingerprints on for a lot of reasons people are excited about but we want to be able to structure on any blockchain so the most fundamental being Bitcoin in other etx o or Universal transaction output that's what we want to be able to do or potentially state-based like some more contracting systems systems like aetherium and hyper with your fabric and others but you can use any of these so we got a builds from ground up the simple systems and then we'd like to be able have a commit reveal process so that you don't even need to expose that you're recording and fingerprinting data to anybody until you'd like to so this information could be as private and secure as you need it to be right and then we'd like to provide options then on top of this so it's an onion effectively if you're exposing that you took a log and then maybe a layer down from that would be the metadata around fat so although this provenance information and then a layer down from that might be actual access to an encrypted data and then beyond that is actually plaintext data right so all of these hierarchy of permissions and controls and access who is to see what we want to build in and this can be at least as we're seeing right now from two different viewpoints the person who is reporting or submitting logs right can be centric around them as an entity or could potentially be around an auditing agency so for instance like a police department might wants to be centered around them asking and requesting for people to submit logs instead of an individual reporter like a police officer submitting a log and all of the system then wrapped an offline first designed so that any clients using corroborator can potentially do this independently of talking to any other service as much as possible and then once they do get back online or potentially back on a mission network we'll talk about you can get a lot more reliability much more assurance that this fingerprint is and the data itself potentially is getting up there in the world so the way we like to kind of colloquial talk will clearly talk about this is Hashem's - right so this is a framework they can work for many applications so hashing is where we can take any arbitrary data it could be literally just a few you know individual characters or maybe a sentence something like this out to video files out to movies out to all kinds of complex file systems that you can put into this algorithm and it'll pop out a unique identifier for that all right so this is a very specific length string of characters and numbers right that will most these systems churn but if you change one tiny little piece of anything on the input the output of that algorithm the hash is totally changed so the use case here is that we can take for instance an image and take the hash of it and then we know that if you change either one pixel on that image it immediately changes the hash completely so you're able to tell if this is the exact original data that someone produced so we take that hash and then stash it we put that on a fingerprinting logbook or database so specifically we're talking about these block chains for the fingerprint and then on top of that if you'd like to store and then share potentially data outside of your device the actual original data were thinking about things like IP FS and other content based address networking's so we'll talk a little more about that but that hash and stature mechanism is a really useful tool and basically is the fundamental building block that corroborator uses for all of this so right now the components then are you take any data that you like and then you hash this and throw that into a time stamping system along with a pro whole bunch of provenance information and then optionally one thing we're really excited to talk about is corroborator x' so other potential P clients who are using this particular application to be able to sign these messages as well to help give credence to that maybe you were in a time in place and that was similar to another one of these people using this client or potentially an auditing service got this and gave you effectively a read receipt that this data was available to them and they were able to read it so all of these things combined really are what we're interested in getting Gordon corroborator so that that kind of describes this universal architecture that we're building and now what an example like and this is really our first in our MVP that we're building is based on aetherium and we'll go into why that is and specifically if you're aiming at the public chain this is our fingerprint system and then ipfs right so we're using this for storage and for our fingerprint generation mechanism and so this is a content based Nate and distributed data store that we can put this on to so this is what we're going forward to begin with as well as then the next steps that we're excited to work on is hybrid online offline mesh so getting other collaborators or using the same application to be able to help both support aetherium transactions and ipfs data sharing so why ipfs so ipfs is this really cool file system it's a interplanetary file system as why PF s stands for that specifically when you put data into the system you'll hash it right and by hashing it that is the name of the file now throughout the network so ipfs is a distributed network that you are able to call out and find files on that work by that hash that's generated so you're able to identify the data and recall it at the same time you're able to validate it once you've got this data so it does both for us and then it also plays very very nicely with the mesh right so underlying IP FS is a whole bunch of other tools lit p2p is one of the things that is built into this that enables peer-to-peer networking pretty simply and pretty sleekly so we're able to do a little bit of both at the same time and then right now we're thinking about ipfs to have a swarm of nodes and by swarm we mean these are pre-established whitelisted people on the ipfs network to share an inter operate data with whether that's just the fingerprints or potentially the data itself as well and then it's also catered to be offline first right we can do this all client-side even on a device as lightweight is a a mobile phone so it's pretty cool we can do everything on whatever device you'd like to bring to us and with plans to expand ipfs into rust there's even integrated devices and things like raspberry PI's and all this that we can talk about very very lightweight devices and why aetherium so as a fingerprinting system we are really keen on using the theory and maintenance because of the use cases that we're targeting hundreds on uptime is very critical here and then extremely high data security so anything we put on to the etherium network it's pretty darn robust that we can access it from somewhere and that the data is never going to change ever all right it's already cross correlated there's a bunch of game theoretic incentives to never let that go away there's a whole bunch of money locked in behind making sure that it never gets lost so we want to plug right into that and it's censorship resistant meaning that it really is hard to block people out of accessing this information as well as being extremely developer friendly so out of the different blockchain solutions that are available this definitely has the most robust tooling that I'm aware of right that we're working with as well as some of the best developer community to be playing with right these guys are generally very friendly very supportive and but also everybody's catching on to this wave so private and permission Ledger's are very easy to hop into so hyper loads are basically for instance is a private extension of the typical etherium client that we can very easily map over whatever we're doing into that private and Commission world if you'd like and they're not the only ones there's plenty of others that are building specifically to be able to transpile with solidity and things like this so it's it's a good good place to start we think so corroborator then we're talking about as a pluggable middleware and the components that we would think that most people want to bring to the table on this are really four different applications fundamental to all of this and really the base level is note right so this is effectively the client that has these core services that is that pluggable middleware component and on top of that we're stacking three different independent players right a relay right which can be a very dumb client this is effectively a white listable client that is just a node that can share information which could just be fingerprints could be some of this chart information and it could be just doing the mesh integrations with these other two applications that are out there so that's the reporter and the auditor reporter really this is an application where you're submitting data could be producing your own logs taking images whatever it might be an auditor this is on the back end where you're able to parse through information and validate information at any time so there's a lot of overlap but in general we think that those are the two primary clients that people will use a relay to help bridge the gaps between everything that sits on top of this node network so the data flow right now our application our MVP looks like this we have a reporter that is iOS app or and also an Android app be able to walk sign and fingerprint all in the same action just taking the picture so you capsules data you package metadata all within the actual JPEG that you're generating for this you hash the file and you sign the file and submit it out to the network right so then a mesh network here can be simply you know you to your ISP it could also be you to another peer over time so we running one of these notes but these peers will push and pull hashes from anyone who's nearby and they can potentially hear corroboree add their own signature onto that original data and then further follow along and push that out to the blockchain so not only is it a single reporter that's issuing this potentially you can ask others in your vicinity or others who you know on the network to also sign on top of this data it's give it a little bit more promise right and then finally we arrive at the auditors side right if you're presented a hash or presented the original file be able to look for the the fingerprinting records along with any of the metadata that if you're giving access or whatever keys you need to look at that you can compare what is the logbook show versus what do I actually have in hand if you don't have the data you're at least able to see some of this metadata if you're given access to that and if you are just passed a file be able to search the log database for if there's a match somewhere in your system so with that I think I'll turn it over to now a video that Ian has put together of a little demo of what we're out with reporter so I'm going to stop sharing my screen there and I'm going to start sharing this video here for everybody to check out first off everybody's still able to hear me yep yep all right sounds good I will go ahead and share this video I'm gonna have to jump ahead a little bit in it alright it's not just us anymore that needs to be the sole savior of it if it goes down so know that I'm gonna hand it over to my colleague mind if that's Ian and Hope go ahead and tell a little bit more about the report or application alrighty this is a demo of the reporter react native that will start an offline mode we've added a bit of metadata as you can see go to the camera and take a picture so or an offline mode we're not publishing hashes to the blockchain so you can see we have a little warning border there we'll go ahead and connect to Wi-Fi and make sure that we're connected here that's good go back to the app and refresh and we published our hash you can see another right there is pending ash that's of EDC DC and then we vetted our chain viewer so now we'll show you that you can switch logbooks and create new ones if you want though we'll just switch to one that we already added so I just have one image in this logbook there's the chain viewer you can see there - dd4l dd4l so those match let's go take another picture paper clip [Applause] that were online we see that it automatically syncs and publishes to the blockchain he started to hash there so let's go to how do we transfer these images pretty simple automatic you just open up Android file transfer we'll change that by last modified so we get the most recent one open up a folder that we can transfer it to let's get that out there so it's as easy as a drag and drop you can see the image is the same of our paper clip now we have an open source tool you can go check the hash this isn't integrated into our other interface yet but it will be seen so here I drag the file over and our hash of 99 100 e you can see it was published on our chain viewer nine-one-one need and then go and edit our file just a little black save it and then recast it that's our stool and you can see that it's no longer nine nine one one that's three C three C go back to our chamber the hatches indeed do not match now that we've added a dimension we can go and check out the metadata as well the inspector you've got GPS data and you can see the image description is equal to that we said in the settings and that's it thank you alright go ahead and stop the YouTube share yeah and I'll echo myself thank you in for go ahead and give us live demo and well I'm gonna jump back into my slides briefly to finish out and then I'll have four questions hmm so so you can kind of see what we're at now and admittedly the demo obviously is a little bit not necessarily user friendly there's still a lot of manual effort that's going on and that's what we're actively working on right now and so our MVP is extremely clean and our first users are going to be potentially first responders and emergency services personnel like police officers so that reporter app is super simple you just click can you able to view your files and potentially do some edits all credit uploads and things are things were working on in the background as well as that auditor interface that right now is very manual and third party is going to be very clean and easy and first party will be providing that for folks so in the next three months yeah number one priority is that audre interface that'll be coming out here probably within the next 15 to 20 days actually and then batch uploads and log submissions is something we're working on actively such that one thing talking to some of the clients that we're working with they don't necessarily want to have every single image be uploaded every time and fingerprinted every time they might want to take some static logs and not to do any of this infrastructure put some edits on metadata or the images themselves and then upload them finally as one bulk submission right and so that'll be coming soon and then metadata and its interface dimensions user interface polish and abstractions will definitely be coming in so that most what you seem to be in advanced mode if you'd like to learn about much transactions and all the hashes and things like this you can but most users will not need to see that all right effectively all you'll need is a checkbox in the end so then in the future right now we're doing images only but arbitrary data is something that we'd like to work with people on figuring out depending on your use case we can help you get that a very simple interface that feels very much like whatever native file application you would use right and then revision dag so more than just the ability to submit single individual files we'd like a distributed a cyclic graph to be able to track provisions along the way so that at any point you can have the ability to make edits to information and reference a previous iteration that was further in the past so that you have provenance for instance the example might be you're at a crime scene taking photos but you need to crop and shade an image and enhance the image but prove to the court that you actually took an original photo that was based on that and you can use that system here to do that kind of thing and then sharded data one thing I'm really excited about in the like forests future is being able to use a confessor and other distributed networks to be able to share not pieces there's give me not entire encrypted blobs of files but pieces of this so I'm PFS actually natively chunks out all the information that you're submitting to it so encrypted blobs and then even potentially further encryption on those blobs themselves you can go crazy with this stuff but being able to pass to a set of peers only pieces the information and no way to reconstruct it themselves becomes much harder for anybody to be able to break into your system right even if they have the keys so this is something we would like to explore and then perceptual hashing so there's some really interesting techniques around like we said the hash is now for unique for an image and if you change one pixel it kills the hash and our we will get fail state and our system that says that you don't pass the check of perceptual hashing and others are designing systems that you can tweak the images in particular ways in the same hash will be generated so this is a little abstract to me and new to me but P hash torque I think has some information and EMI people will speak to that a little bit more after if you guys are want to learn more in the more plugins so right now we're talking about hearing we're talking about high DFS but we'll bring your own what whatever you'd like we'd like to be able to configure this for whatever you'd like to submit to depending on your use case and then more fine-grained permission right as I was looking to you can get all kinds of onion wires so we'd like to explore what clients would like so with that I think we can close out just putting this thought in your head is that really this system we think if we design it right is effectively as good as you being there in person observing this event all right can give you that same level of a surety that it's tamper proof gives you an ability to effectively watch the watchman as well and maybe this is a point I didn't stress hard enough is that even the people who are operating and administrating the system can't tamper with these fingerprints and these laws so it's really down to the individual level that you can have trust so we can effectively have audit systems that watch the watchman right and anything can be corroborated right so whether this is images for police officers or potentially guarding against fake news and all this stuff and deep fakes somewhere exploring so I'll close it up there and ask you all for if got any questions and thank you all very much for coming down are you doing this as the for-profit business or what yeah so right now we're exploring what our product market fit is specifically and we have some pretty hot leads in the really case management side of things so whether it's all the way down from the police officer all the way out to court that whole chain of command for this data it's something we're exploring like there's systems that we think we can compete with pretty heavily so for profit yes but where are cuts acting is open-source right so the open-source repo I can share it's in the slide deck and I can share it to anybody who's interested afterwards but this pluggable middleware we'd like to let anybody use and ideally the the profit model will come from helping people navigate that customizability because it's not obvious to how to structure this thing you might have a great tool but it's a bit of a you know i what's what's the Swiss Army knife right we'd like to teach you how to use it and give you the best tools to use for your particular use case so I think that's what we're really going after so you'd like build a phone app for the Denver Police Department so they can use it that kind of thing yeah so we can build potentially whitelisted apps for you or potentially give you this API interface to plug into stuff that you're already using so this can be an add-on to systems you have but we can help you configure like which blockchain do you want to use potentially if we're talking about a cryptocurrency based system like aetherium then we you might not want to deal with gas so we might be your gas provider all right so you'll have to touch any of that stuff so it just really is client specific what they need okay thanks good presentation Thanks thank you looks like we have maybe a couple things in the chat huh yeah we have at least one person who works in the courts and record system they will have to definitely talk more about that once we get your mic to work yeah and I think Ian also posted in the chat corroborator the open source files are all on github for you so it's corroborator - net what you want fact I'll just bring it up and share my screen while I've got you all right so for anybody who's looking you can see this is a stale document at the moment but will be opened back up it's probably the best place to start is this overview so we're dig into this and give a little more of a status on where we're going where we're headed but there's some more videos and presentations to check out if interested as well as some of our design choices and some of the things that we're using in the back end and yeah so check us out github.com slash corroborator - net no worries well will sync up later definitely we have any other questions Allen or you know drew no questions uh can presentation though good work thanks thanks all a great development work here you can think and primarily for putting together a lot of these ideas and making them a reality I'm definitely not as heavy on the development side as you are I I'm gonna catch up again but so what I would ask everybody who is watching is that if you know of people who would like to use this kind of system please reach out to us we very much like to get in touch so I'm Dan at crypto rhetoric you can also find me Newcombe and Dan Anki bass is a great place to find me or on Twitter those are the great places to get me so we'd very much like to talk to you about whatever the use case you might have that we can get your help on as well as if you have ideas to connect us to others who might be on gives of themselves or suggestions improvements what what are we doing wrong please please reach out all right thanks again Dan thank you guys and I'll post this reporting here shortly for anybody who wants to review thanks Joe thanks