 Hello, everyone. I am Linnea Nelson, ecosystem manager of OpenIDL, and I will be moderating the discussion today, and I will kick it off to Joan to introduce herself. Hello, I'm Joan Zerkovich and I'm the Senior Vice President of Operations for American Association of Insurance Services, and I also serve as a technology lead in the organization. In that role, I am the executive sponsor of OpenIDL, a project under the Linux Foundation. Great. Tram, if you want to introduce yourself. Hi, everyone. My name is Trambo. I am CEO and co-founder of Mobi, and my background is not a technologist at all. So I think the non-technologist on this panel is actually was in chemistry, and then I went on to art conservation and then got into blockchain, got excited, look at it from the intellectual property management of that. And of course that leads to mobility and smart city because we all in a digital age produce content. And in 2018, I'll launch Mobi along with my co-founder, Chris Ballinger. Great. Good afternoon, everyone. My name is Ryan Briggs. I'm a part of SwissPrix Solutions, and I lead a function called Automotive and Mobility Solutions for North America within SwissPrix. We focus on technologies, data products, and consulting services that help our partners in the insurance industry better price, risk, and create innovative products. Awesome. And James, last but not least. Yeah. James Madison. I've been at the Hartford for 21 years. I'm an architect in the data office, and I'm on the governing board of Open Ideal. Awesome. Well, I want to thank all of the attendees here today. There should be a great discussion with our panelists. I'm excited to learn from them. So we'll kick it off. Blockchain hit the insurance industry by storm about five to seven years ago, promising to solve all of its problems. And its reputation has dropped as it did not solve all of the industry's problems. But we're in a great position now, knowing one, that the technology works and two, where the technology best fits as a solution. So it's no longer a solution looking for a problem. It's actually solving problems. So we'll jump right in and see what today's panelists have to say about blockchain's evolution within the industry. So the first question I have, panelists, can you describe your perspective on the life so far, a blockchain in the industry and where, after it's ups and downs, it best fits as a solution to enterprise level industry problems in the areas within the industry that it best fits? Well, I'm happy to jump in and start the conversation. I think blockchain within the industry suffered as a lot of trials with the use of the technology did from the confusion around Bitcoin, which is one way to use the technology versus the many other ways that the technology could be used to solve real business problems. And with that, there were a number of consortia that came together to see how blockchain could solve some problems that we had with the exchange of information and data. And in those trials, the technology always worked. That really wasn't a problem. There were a lot of business models around the use of the technology. So we do know that the application of blockchain technology to keeping data secure and private and immutable records, all of that works. But how do you form a network that can take advantage of that technology? Like it's a single application that you can pull off the shelf and install and you're finished. It requires businesses to come together to form a network to agree how to operate using this technology. And I think that's really where we've struggled with the success is the business model, not the technology. I couldn't agree more. I couldn't agree more. Even though Moby, when we launched in 2018, our history went back a couple years, and it started out. Many of them now Moby numbers were very excited about blockchain and the potentials with other issues that the industry were having data privacy and connectivity and all that. What I found out is that when they do a POC, a small scale POC and they can put a vehicle on chain service on change, all that was possible. But there's the application couldn't scale is because everybody keep using decentralized technology and building centralized platform. For example, BMW spent a lot of money building a supply chain platform, but couldn't convince other car makers to go on to it because it belongs to the other. And Daimler Mercedes did the same thing for mass multimodal and DHL did the same thing for logistics. And they waste millions and millions of dollars. And due to that, that's why we found it Moby so that we can create standards, standards like how to identify vehicles things people a trip. When does a trip begin when does it end that needs to be also identified how to settle business transaction, those needed to be standardized. And then the second thing is how do we offer business automation for the ecosystem using decentralized ownership and centralized platform and build that out. And decentralized is the key there I agree with you, Tram that a lot of these models had a small number of participants and they were still working in that centralized model and inherent in a blockchain network is a decentralized model, it's a network. And I think that's where we're really starting to have some success now is now that we understand that through all these trials, we're now moving in the right direction. Yeah, really quickly just for our audience, who may not be, you know, as well verse in, you know, the vernacular that we are used to. Can someone define, you know, the difference between centralized a true centralized and a true decentralized blockchain network. James I don't know you want to That's fine. I'll keep it simple. No, I mean, blockchain the technology is extremely well proven. What you're seeing is some of the silliness of the use cases that came and went around it. Maybe some are still going. No, it's the idea that everybody gets a copy of the ledger and everybody can verify what everybody's done in the past and nobody can deny what they've done. So that we all have it and that nobody's allowed to have 51% ownership because you know that kind of thing. So it's just the idea that we all know what we've all done and nobody can deny it and such that everybody has a copy of it, as opposed to some trusted authority it's like well I believe the bank because the bank is reliable may have been around 150 years or whatever that's a more of a centralized model so The notion that it's decentralized that all the peers have to agree to some degree, you know, without getting to nuances is the key, we all have a copy, we all trust it because everybody else is trusting it and if we don't trust their peers then the whole thing shuts down anyway that's it has to do with the technology. The thing about the blockchain is it's a question of the use case. And so the open idea of the use case that the heart is most concerned about is being able to have extremely high security and control over our data. And so if you can get this model to work, it provides that. And I say if simply because this is a massive effort right so it's not if you're in terms of the technology or the idea is the execution and the number of people you have to bring to bear on and the number of standards you have to publish. So any any hesitation I might seem to have is strictly around execution and just a sheer effort to get this done as opposed to the design itself the design itself is elegant, which for me simplifies down to instead of sending data to the query you send query to the data. That's the fundamental difference right so we as a company are not comfortable sending our data out anywhere more than we have to. And so there's a very simple principle of computer science that says you can send data to some place, and then that place can query it all they want well the problem is they might query more than you want right so we don't want to send our data anywhere. So we want to keep our data here and you say no you send me the query. And if I like your query, I'll let it execute, and I'll give you the results and those results better be very rolled up so you can't detect things I don't want you to detect. So that's the fundamental shift so that business case is what we want that theoretically and computer science goes back decades right probably to the 50s for all I know, but it's like you want to get there, but getting there's really hard. It's actually pretty simple to just hurl your data to somebody else and like an insurance officer or whatever right and just let them do their thing. It's really hard to structure your data in a way that allows you to bring in queries, get only the results that back and send it, but doing that and then doing that on the blockchain is we're interested mostly because it just it lets us keep our data and we say yeah I don't like your query I'm not letting that go in. Oh I do like that query. I will let that one do it. That's the whole the gating of all those queries is also a key design principle with that too. Awesome. Thank you that was great and Ryan, you know, from your perspective, you know, having this high level view of the industry and bringing and creating new solutions for your clients for carriers. Where do you see, you know, the decentralized model being, you know, essential to your work, your vision. Yeah, thank you for the question. Yes, I think blockchain and distributed data sets have an opportunity to really stimulate innovation. There's so much innovation, especially in the areas that I focus in in the automotive space, where there's new technologies, new business models that didn't exist previously. So it's really cannot wait for five years to develop last history on any given one of those things because you'll be constantly waiting. It won't ever happen. So, and really the pace of vehicle safety technology for one specific example is far outpacing the rating models that are prepared to use there. I think in stimulating innovation is a key use case for a distributed model such as this. And it was occurring to me in preparation for this, that there's a behavior that has a promise to change when it first, you know, by nature don't have data to understand a risk or don't understand the risk. The behavior is actually to withdraw from it. And that is really a promise that we need to reverse. So instead of withdrawing from that risk to use a collaborative approach to understand it better, and perhaps look for data sources, a you know, that can better inform that risk. And so you can attack it and address it for the good of the insurance and society. That's the behavior that I would like to see that come from this. Absolutely. Thank you. So as far as, you know, blockchain and we covered this a little bit. But as far as blockchain models go here we're referring to solutions built on an open source decentralized permission based blockchain network held together by an open governance model. You know, panelists, why do you think this is the best model suited to solve enterprise level industry problem problems. You know, and maybe not to dive too deeply but we can use examples of where previous models or consortia have not worked and maybe describing why this model is landing landing in the sweet spot. Yeah, yeah. So, first look at the model of the internet. Why do we use it why do all businesses use it. We use it because it's open source. It's not controlled by a single vendor, a single group, it is open, and it's accessible, any company small to large can join and participate in the network in web based services. They can afford it, it's accessible. And, you know, going back to what Ryan mentioned it provides a place for innovation so if you think about extending that model to a problem that we've encountered while we've started to use the internet that's great we love that but we're still struggling with data privacy and security, trusting it, centralization, those things didn't go away and so that's a problem we have to solve. So when we go to solve that problem we have to go back to those core principles again which is open. Building a decentralized model not centralized decentralized that enables everyone to come and participate and control the destiny of their data right we control where it's held it's still held in house just like we do today with any information we provide over the web. We hold that information in house and we control where information goes or how it's exchanged and so if you think about the networks that are developing to solve that problem and in this case blockchain has proven to solve a lot of the problems with data security and privacy and all of that. We're going to hold to those core principles. So the source decentralized allow the endpoints to control what's happening to their data. To me that that's really critical and the other part of the question is, you know where have blockchain initiatives struggled in the past I think it's, again forgetting those principles that very a lot of these initiatives were initiated under large players in an industry right because they had the resources the staff to go test the possibilities, but they didn't have the benefit of what it meant to the entire industry the smallest players those that don't have the technical specification that maybe the early participants do, and you know they can afford a 250,000 outlay for software and then you join a consortia and then you run this once you hire the staff to manage it. And that's, that's where the struggles have come in where we've tried to create these kind of centralized, somewhat proprietary models and so again if you want the success we have to start simply on an open source solution, make it very easy for all the players to enter and then see what should happen on top of that. Let the carriers. Now that they have the tools very simple set of tools just like the very simple web services that initially that initially were available. They'll find the best way to build on top of that network. Absolutely. Anyone else tram do you have. Yes, I couldn't couldn't say it better. We both, we both are from consortia point of view. From the beginning, we, if you look at any industry that's hundreds of thousands of ecosystem stakeholders. Public their private, they all have unique databases they all have unique processes, they all have different regulation depending where they they are for handling business processes and handling customer PII. So if you have these hundreds and thousands of if not millions of different requirements. How do you have business automation happens for all these stakeholders. And I think they have five key things that we need to keep in mind and some of them are repetitive of what we've been saying is, first is zero trust authentication. So that means every entity must be identified and authenticated for every transaction. And that any data credentials or claims must be non-recruitable. I have problems with that word. The second thing is data privacy. That means regulatory compliant. In Europe and now in the US many states are having that now. And then you have to have limit access to intended recipient only so you're not opening up your database to anybody who wants to come and do business with you. The third thing is data security and selective disclosure. That means you don't want to connect to centralized databases to do business. James was saying you want to push the data to the edge and do business at the edge. And then you want to be able to selectively disclose or verify information at the moment of transaction and not have to store the data before, during or after nobody collecting it. And then the solution, as been said, it has to be affordable, interoperable, scalable and extendable. So to meet all those requirements, we have to use standards. There has to be standard space. It has to be platform agnostic universal translator as we call them. That works with any legacy system that could work with a mom and pops little store that makes a little nuts and bolts for a vehicle and a big company that makes batteries. And they don't have to build new infrastructure or continually maintain the infrastructure or integrate with somebody else's platform. And then decentralized ownership. It has to be community owned and operated. And I think those five, those five things, I think books, then business automation happen. Everybody can come and play. Yeah, I can build on that too. Because you hit one with really good language. I love that selective disclosure at the transaction level. You said the selective disclosure, I read it a little bit, but the idea of we can approve every individual query. So one of the challenges you have traditionally with data is that if you give people rights to the data, they can query in any number of ways, right? But if you have the ability to say on a per transaction basis, yes, I'll let this win and no, I won't. That selective disclosure on a per transaction basis is absolutely critical to the entire design. So I love that point. Two things that matter for me too is the notion of the standards, right? Standards particularly around the data model. So in principle, this whole network can do things besides insurance once at scale, right? But I'm an insurance company. So my question is insurance. But if you're going to send the data to the query, excuse me, query to the data, let me get that right, right? The data has to have a certain predictable shape. So if I want to fire a query out to 100 carriers or whatever, I just keep using insurance, right? If I fire this out, this query out, the data structure has to be in a certain shape. I need to expect to find a vehicle object when I need to find it. The third attribute on that thing better be number of doors, right? So you do have to have a common data structure in order to do this. So one of the biggest things I'm watching for in contributing to as much as I can is the data model. So that's consistent and that's going down a good path. And then the other thing I do observe is the openness because the one thing that no vendor product can compete with is open. And so there are vendor products that have some of this functionality, sometimes a lot of this functionality, right? So the idea of sending the query to the data instead of the data query, the idea of selective disclosure, et cetera, those are standard principles. Anybody can go out there and code them. It's in many vendors self-interest to do it. And you can find these functionalities in vendor platforms, but they're always going to be proprietary. So it may take a while for open source to go. I mean, unless I love open source, but it's like, yeah, it's like, I don't know. It's a lot of opinions and a lot of things that you have to kind of work together. So it may be a little slower to try to get a whole world community onto an open source project, but it will always be open. Whereas vendor product, yeah, if they want to, they could pay 100 engineers to slam that thing out the door and have it in the next product release, but they're always going to be closed. So to me, that's one of the biggest steps for engineers to is that it is inherently open. And so it may take some time to get those functionality to fully function, but it will always be open and that's key. And I think just to add to that, one of the areas that OpenIDL felt was important right from the beginning was to include the regulators. Because all of these solutions that are being developed, they are being developed for an industry. And the role of the regulators is in part to make sure that an industry operates in a fair and transparent manner. And they need to be able to understand how that industry is operating. So if it's closed and they can't understand how that technology is working or whether it's discriminatory to somebody in that uses the products or even to a group of businesses in the industry, that's a struggle for the regulators. So they really need to understand any of these networks that are developing and OpenIDL from day one has worked collaboratively with the regulators so that they understand open, right? They weren't connected to the open community, so they've learned a lot about it as our participants have. And they're very excited about it because it meets their need of that transparency into what's going on. Absolutely. And, you know, with that, you know, I want to toss it over to Ryan and maybe, you know, on the note of regulators and working with them. You know, Ryan, where do you see, you know, the possibility of matching, you know, in auto, our reading system to the technology and where that and, you know, how the regulators and why they will be so important in that process as well. Yeah, I think there's a there's a parallel actually with with regulators because regulators have domain over a certain, you know, political geography, and they want a functioning market within that place. All right. Reinsurance actually has, I don't think it's a stricter analogy, a similar function where we as a reinsurer have domain globally and the reason for that is that it's distributing risk transfer from insurance companies but operating globally so that we can balance risk in that way. So, when you have a what a what a regulator for specific specific political geography is looking for is market non performance in their place whether it be behavior or the worst case is risks that can't be covered because then those insurers are in jeopardy. And we see, we see behaviors like this like in my space this is one area as an industry commercial motor risk has been unprofitable as an industry for going on I think like 15 years now, right. So, so all, you know, some dedicated line insurers have gone out of business. Many are just subsidizing it in other places. Why hasn't this problem been solved. Right. Why, you know, what, what's, what's going on with that. So within a given, you know, market. What's happening is insurers do what they do. They are trying to write better good risks and less bad risks and the, the, what happens is the bad rest either go to another carrier who tries to do the same thing, or they those bad risk a lot of business, right. And in some ways, especially when the rest of us tear the same roads and our livelihood, you know, and literally life is that jeopardy based on the risk of other people operating. I don't think it's actually serving the market well to have from an insurance perspective or from a desirable perspective. So, I see opportunities such as, you know, within the technologies that are just discussing and I'm not getting into it to that. I think that's the opportunity for the opportunity for collaboration to really resolve different risks used across insurance carrier platform, such as, you know, centered around a regulatory opportunity or centered around a risk capacity opportunity to try to get out those problems and solve them for the greater good. Absolutely. That's beautifully said. And, you know, I'd like to just, you know, illuminate the elephant in the room as well and talk about the challenges. In the past, the challenges we're facing now, but also, you know, once we overcome these challenges, the opportunities that we do have. So I'll let Jonah let you pick this one off too. Yes, there are always challenges, building a new system. And certainly as James mentioned earlier, building a network is more of a challenge than a vendor supplied solution that you can just go by off the shelf and and implement it so that's the challenge and the challenge of bringing these new solutions in. It does require data standards, and that takes some time, but I think that the challenge to the groups that are doing is keep it simple, right start very small to find those standards in a way that you can grow them over and get members into the network start building that that network over time, iteratively, so that you understand the value of the standards the data that's being used the solutions that are being built on top of that. So when you see a big problem like that it's best to kind of break it up and and make progress a little bit at a time. And, you know, the example that I give most often is for folks to go back and remember at the beginning of the internet if you're, if you're around at that time. There were centralized solutions there were vendors that provided network capabilities on their single platform, and they were all offering solutions at the time the internet was being built and how was it built. It was built through the collaboration of industry with the National Science Foundation and researchers all over the world, and it took many years to develop it. And they developed the network and then some physicists came up with this idea of exchanging documents across the network using this web thing that they thought was a good idea. If you were to look back way back then you would say I'm building this network is really big and it's really hard and, and I'm not really sure if it's going to be useful and they have this crazy idea to exchange documents across this network but but because it was open. And people were building it iteratively. The physicists started exchanging documents, and they started developing some standards around how information is exchanged and how this web thing was going to be used and I and I remember where they said yeah you connect going to HTTP colon slash slash and everybody said, Oh, nobody will ever use that that's just too hard. They said well it's what we've got so they started building web platforms and today we're still typing HTTP colon it wasn't the usefulness of the service so great that as we've done it over time. We never saw in the beginning that you would be conducting banking over a web based service and so that's when I think about this is hard building a network is hard building data standards is hard. Yes, but we've done it before, and it resulted in the best solution and we're doing it now we just need to put one foot in front of the other, solving the problem every day. Yeah, it's a great example. Yeah, yeah, yeah. So four things that I can think of that worry me right so one I mentioned before is the data model and first and foremost a data person right it's like you know I have a computer science master so I do technology for fun unfortunately I get paid to do data because data is more important. So I know I just made all the software developers out there mad. But to me the data model is the key right we got to get the data structure down to the right thing for better or worse let's just take something like the stat model has been around forever the stat model is in fact a well defined consistent format that we all use. It just has, you know, all manner of problems. So we have to get just a much more robust model and I kind of joked that the current model for just for staff violence is basically a 1970 model. Literally, and we squiggly braces to me negative numbers and weird stuff like that. And if if you ask a question of it and is wrong you submit a request you get another six weeks, you'll get another report and you'll ask another question you wait another six weeks it's literally a 1970s paradigm. We're trying to get up to just something even as good as just late 90s or early 20s with the slice and dice kind of interface that we're all used to that every kid out of college can now do in a pivot table. You got to get your data up to that so we're not even like shooting for rocket science forget like you know big data and munging and everything else it's like. Look, can I just get up to like the year 2000, where I can slice and dice and that sort of thing so that's powerful, but to do that you have to have the data model structure to correctly got to get your grain right and all that good stuff so one is the data model and getting a whole army of people carriers and regulators to agree on that stuff. The other one is the go no go gates, you're talking about plugging your data into a network. That's that's scary it better be really safe not just safe technologically like the crypto is all safe that's fine the blockchain safe. But at the application layer you have to have gates so one of the things we want is when you bring a query into our environment. I want to stop that what is that query. Do I trust it. Okay I think I trust it let me run it. Oh wait a minute. Let me look at the data before that data leaves to so while it will be on the network. There's got to be stops every step of the way and if we say yep I want to check every step away needs to be there so we're building in those go no go gates I call them. The other thing is you got to have participation counts and percentage level so if we are trying to contribute data to some kind of environment and it's going to be anonymized in some way by by through aggregation. It's just like when you do like employee surveys you know if you're a manager you only have five people you don't get to see the individual results if you have 100 people you do. We have to make sure some of those safeguards are there too so that the participants are inherently anonymizing so we're watching that as well. And the other one that's interesting is going to be what I call fuzzy hashing so the way that technology works is through a lot of hashing so we're going to we're going to join things in a way that you can't see them. So you can you can use the data without seeing the data is the way I put it simply right off for the geeks out there you hash the data and the hash is still joined even if that you don't see the data. Well the challenge comes when you have fuzzy things like addresses and names and phones. If you don't master that in some consistent way it's really hard to get that stuff to join anonymously. So now let's face it that's hard inside the company to when you have full control of your own data so you know it's not like it's somehow more difficult it's just that you know it's an inherent problem. So watching out for how do you do like you really want to get into some of the advanced stuff like we know we can do staff filing over this network gay we know we can enhance the model and do something way better than that by gay. We know that we can do data sharing and any parties who are interested in sharing as long as the data is crisp and it can be hash cleanly okay. It's just when you get to the fuzzy levels going to be interesting but now we're like like crazy advanced levels but those are the things that I'm just keeping an eye on. Yeah. Oh, yes, thank you for movie I think some of the most difficult things on the hurdles that we need to go over is versus education. This is a new way of doing things. So how we educate our members and the general public and get more companies to come in and join is education is a new way of doing things a new language, learn the alphabet before you can write poetry. That's number one. Number two. There's different level of understanding and at different times throughout the whole ecosystem. So there's early adopter and then there's late adopter and how do you make sure that the early adopters still stay engaged waiting for the late adopter to come. That is also another hurdle that we are having. And then, of course, the data model like how do we agree. Even like for example we call them battery birth certificate. What, what goes into that we all have to agree before we can actually start communicating if this battery is really who it is. What goes into that data. So, if this system works. I think, you know, open IDL, the ITN, which is what movies doing and the dystopia will stay around for a long time because we need all these data standards. This is a really new way of doing things. Somebody needs to do those. Absolutely. And if I could, we said if a number of times right so let me address that because to me you shouldn't have any fear of the blockchain that's just that's that's good tech. The theoretical pinnings of this idea. And again, I joke I'm a theoretical computer science but I scientist but I work for a living because I have to pay my mortgage right. And if I geek out too much stop me right but the theoretical underpinnings of this whole move the data, the query query the data, blah, blah, blah. That's all sound fuzzies hashes all that good stuff right. If you buy into the theoretical stuff or just believe some geek like me. Okay, cool you have several staff asked them they'll understand what I'm saying right. And you believe the blockchain. The question becomes execution like and I will say the question to ask is you know do we have the right bunch of people on this thing and by the way it's all public. It's because if you anybody can join the meetings that they want to. But if you come into some of our, I think we have three working sessions right now and not including all the governance stuff. One around the model one around the infrastructure and one around the architecture of three major ones I know of there might be more. We really have a good army of people on this so part of the question you want to ask is, if you're going to pull this off who do you need and I think we have them. So, some of these meetings we have 1015 20 people on them, and they're people with with 3040 years of experience who in the insurance insurance industry so you know if anybody cares to jump on to and talk about you know what we have behind this we have the right resources. We want more. So if you're listening and you're interested please join, but I'm very impressed with the people we have and it's really fun to just like design the stuff and think it through. So there is also the question, can you believe in the execution that we have going on, and I do. And I think we got the right bunch of people on it, in addition to the right theory in addition to the right technology so it's good stuff. It's just a long road. There's a lot to do here. Yeah, absolutely. And Ryan, where do you see, you know, from your unique outlook in this, you know, on this panel, where do you see, you know, the biggest challenges or, you know hurdles in the blockchain space. In the industry. So when I look at this, I'm not going to be the one that speaks about the technology on this call. I'll leave it to Tram and Tram and James who can get much more into the science of this so I focus on the use cases, the really the business opportunities that this could help resolve. I see them. Most of what I immediately go to is where the problems in the insurance industry are right where I see within my own organization or hear it from other organizations where they're like, well I don't have the data so I can't seem to do this, or, you know, we feel like there's a way out there but I don't have the data, so I don't know how to do this we were being pressured from regulators to not use that data but what data do I do what data do I use you know, then there's people have literally have business models that keep data in, you know, behind a closed door and you can monetize it right so that's what they're doing and in a way I get to this you know kind of word where it becomes a non functioning marketplace where there's everybody putting up barriers to working with one another. Right. And that's the business model is to make it difficult to work with one another. So I look at the, the opportunities here of where we can come together to, to solve these problems and yes, make a business out of that as well. So, you know, in use cases that I, you know, personally involved in so these may not be the biggest most important, you know, the title or business opportunities in the world is someone that Ryan cares about. So, but just look at your vehicle safety you know how, how do we prove out the frequency and severity benefits of vehicle technology that costs thousands of dollars. Right. What, how do we prove out which, which way is, is, which system is better. How should we address the risk of electrification. How should we address the risk of care mobility care mobility I would say argue with another one was a non functioning marketplace, because the insurance costs are ridiculous that the barriers are coming in and out of it, and therefore knowledge is great knowledge is lost in, and when it enters a market and leave a market. So, anyway, that's, that's where I focus is these opportunities where everybody has struggled, succeeded and failed and blockchain has the opportunity to least create some persistent knowledge that ultimately would yield to a better performing market. Absolutely. And, you know, to that point and I'm so glad you brought up use cases, I would love to go through, you know, current use cases, POC is past and successful ones that we've seen. So, Joan, I'll let you kick it off with that one too. Yeah, and I wanted to follow up, you know, Ryan you bring up a really good point in that you're interested in information that is inside a network being built around mobility and the auto industry, you need information there. You need insurance information in order to ensure a lot of the products and services that they would like to offer. So I want to go back and point to networks are critical to that a centralized application, it's you're just not going to get there with that approach you have to enable the information to be held by the carriers by the automakers by the mobile service providers and you have to have a way to exchange that information and time in. So, as I said, each one of you are you're going to hold on to your information out at the end points you're going to see that query and you're going to decide how you want to answer it whether it's to an automaker or whether it's to a carrier you have to have a way to do that some standards to do that. And by having an open network, you can do that the centralized model breaks down very quickly if that's the way you're thinking about it. We recently had a proof of concept that we ran with the North Dakota Department of Insurance, it was sponsored by Commissioner Godfrey, and he wanted to know how many uninsured motorists there are in North Dakota. And he has a really hard time getting that information. And certainly it's the statistical data that we send today is two years old by the time a state gets it so he doesn't know how many uninsured motorists there are today because the data he has is at least two years old. And so he wanted a new way where he could ask carriers, information about car, we have cars that are registered in North Dakota, are they insured. So that it's, you know, do we have a new technical platform that would enable me to ask that question to get timely information and do it in a way that works well for the industry and he said up front that he wanted to collaborate with carriers to see if there was a new technical platform that would enable that. So the North Dakota proof of concept was to prove that blockchain technology works and the and the commissioner after the end of the proof of concepts. Yeah, the technology works that there's no doubt about that he and, but he said it was interesting what we found. So with the queries that were on the network, we were able to match again going to James's example is that we had hashed vehicle identification numbers that the carriers have we have similar information a standard to data standards we have vehicle identification numbers that the Department of Motor Vehicles has, and we were able to match those up never exchanging the raw data or the vehicle identification number it was never exchanged never out in the open, but we were able to say is this vehicle that we have registered in North Dakota does does it match is it a policy in force with any of the top 10 carriers that participated in this project. A couple of interesting things happened there the first queries that we put together showed that 20% of the vehicle identification numbers didn't match but not because they weren't insured data quality data standards is really what came to the forefront there we had malformed vehicle identification numbers we had some records that were ensuring an ATV and it's not really a vehicle that's going to be on the road or with the Department of Motor Vehicles. So it wasn't the POC was not about it was to prove the technology, it did that. But this is where I mentioned the iterative approach. What it did surface is that we need data standards, and we need a way to help the manufacturers get the best information to the state so they can make good policy that benefits the industry as a whole around uninsured motorists. So it was a it was a great proof of concept, and that that wrap up is going on right now. And I wish I could tell you what the follow on to that is, but the commissioner in his message to the group said, he wants to continue in this open collaborative environment in which we build a network said the technology works. His approach though is to continue, including the carriers with the regulators to develop the solution and and I think that that's critical success. Absolutely. And it's, it's a great use case and I think specifically for that one Joan it really the weight of it was what you know we didn't know, and what we found within the POC and I think, you know, to that point, there's so much more that we will continue to find specific POCs and it is almost an R&D play and a lot of ways for everyone that wants to participate. So, I want to open it up, you know, we do have 15 more minutes I want to say to the attendees here if you have any questions for our panelists, please put those in the Q&A box. And I can run them by everyone but going back to the current use cases, Tram, if you want to, you know, discuss any of those with Moby, that would be great. So I think the key to what we call the new economy of movement is the ability to link a trusted identity with location into a verifiable trip. And, but we know location is PII, identities PII. And so how do you do that in a safe way that you can still trust the location and trust the identity, but not revealing the VIN number the driver and all that. So one of the use case that we just finished a pilot with our members is the dealer floor plan audit. We all know car dealership within the US and car dealership finance their loan to be able to purchase these vehicles to be able to sell it to the dealership. And the lenders to the dealership, they want to know if their collateral is still on the lab. And there's maybe a few dealership sometimes they sell the vehicle and not report it because they want to hold on to the money and do more things with it, instead of returning the loan right away. And so how do you keep track of all these vehicles are they still on the dealership, the lenders have to send physically somebody to the dealership with a clipboard, sometimes with a scanner and still have to check off vehicles on the lab. And that's a really time consuming way of doing this, this, this audit. And within the US alone, if you can automate this process, it would say four to $500 million, and that can trickle down to the consumer. So, in the pilot, we have the lenders pinging the vehicle and asking the vehicle, are you in this geofence location and the geofence location is the dealer lab. And then the vehicle then in turn, just answer yes or no. And if they, they, yes, they in the dealership and if you ping it for 24 hours, once an hour, and consistently, most of the time the vehicles yes, yes, especially at 3am at night right that vehicle, that'll be in the deal a lot. So, then, then you find your dealership has all those vehicles in a lot. But if it's consistently ping say no I'm not. You don't need to know where it is it's just not in the dealership it could be out driven by a customer. You don't need to know that location. And if it does it multiple times you have issues then then the lender needs to decide what to do about that. So an automation like that could save a lot of money, a lot of time, and near real time data can then that's foundational to many use cases to build on top of it. Like insurance for example, most dealership now do not have health insurance health insurance at all because that a lot of dealership don't have a cover. And then how do you guarantee that the vehicle is actually being stored under the cover when the hailstorm can. So I think knowing this kind of information would be very helpful in the insurance industry. Absolutely and I think that use case to tram, I, we can all everyone, you know, in the industry can place, you know, that use case into many different use cases in the industry. It's fascinating and it's exciting. James, do you have. Yeah, go ahead. I was just going to say it to me what tram just talked about ties into what Ryan would like to see Ryan wants to know what's happening with that electric, electrified vehicle what battery is installed or it has the latest software installed in this car and the insurance industry needs to be able to get access to this information but the auto industry is going to be very cautious about exchanging information because very often it's personally identifiable information and so that we stop it just as Ryan said we stop working together because it's hard. But if open IDL and Moby are working together now Ryan has a way to ask is the latest version of the software in that electric electrified or auto driving vehicle. And because we've been working with Moby Moby only has to come back and say yes, and Ryan trust that answer. That's the power of having it built in a network in which we have data standards and we can talk to one another. Yeah, I'll have some color to that in a few months but you know I've been working with open IDL and you know the, you know had a few calls with tram now as well. We have more use cases are keep coming into mind. You know your tram just use the dealership use case of really just infrequently knowing about when a vehicle is at a dealership or not. We're within, you know, our business of course, where a vehicle is when a storm hits is a super big deal for property and and vehicle insurance. There are also big governmental and societal issues around fuel use tax and tolls ways, new models road uses purging and things like that that are for models of just for example throwing out use cases for people to to to feed on, you know, electrification is going to have a big impact on who pays for the roads when you're not buying gas anymore. Right. So, you know how how you do that are big issues and there's opportunities to do that. I would, I would spend this a little bit maybe throw a softball over to James is some of these use cases seem to be not that complicated. Like the question don't ask me. You don't have to expose a rich data set that needs to be queried. It could be a relatively simple data set and therefore one thinks data standards would not be that big a deal. But I don't know. There's my softball to you James. Yeah, there you go. Yeah, there's a distribution here right so you know you typically especially for engineers like me we have a problem which is we jump to the hardest case. Because it's because someone's going to ask us to solve it right, but you got to sometimes remember that you know most problems are 80% easy 15% medium 5% hard. And so when you get into like fuzzy hashing which I'll explain as I go here. You're up to the 5% use case you're right there's a whole bunch of bread and butter at the bottom it just works. And so yeah, we would have to keep in mind that the distribution of complexity on this is there's a lot of simple use cases we can just win on we start getting people plugged in. A couple things to observe right. So if I go up in scale to I don't want to talk about any specific use cases that gets a little bit competitive but the one of my other responsibilities is third party data for the institution and we bring in hundreds of data sets like all companies do. But what's interesting is even though one of the major first cases that we're looking at this is in carriers feeding regulators. And we're looking at the baby steps first right again start simple which is just can we just do stat this way okay then can we expand that okay. But what's interesting is you start to see opportunities where it's more of a b2b type thing where if other organizations get on to the network. We can start sharing data that way. Now granted we already do that all of us have ridiculous number of fees coming in out of organization but it's everyone's a different story and so on. If you can get this consistent we can you can find is that organizations whose job is data start to want to pile on trans organization kind of fits into that category already. But even all the traditional third party data vendors start to become maybe interested parties and plug in here and you can already see this there's a number of like whatever you call it data marketplace type vendors out there and in order for these data marketplace type vendors to be successful they have to have enough of the big data players lowercase be big data players plugged into their network. So that plug into that network is key but if this is open and people start plugging into the network. And then you start having opportunities beyond just carrier to regulator because well as the first line of thinking just like Joan said the first thoughts of how the Internet would work was not that your banking would be on there. If you can get there, it then becomes the interest of businesses to plug themselves into this now we can start doing more exchanges along those sorts of lines. So that kind of that's an entire category of use case which is just all the data augmentation you want to do you can start getting on the network, particularly if you have fine grain security again at the transaction level. And that is, we keep talking about zero trust, because in zero trust, meaning that even if even if we don't trust each other we can still use the pieces of data we care to share and we'll both be successful as a result and yet we'll both be safe right. But there is something to be said for one way trust relationships, because if you're going to start piling a third party data onto this network if one party can trust the other party in other words if I'm a 3 pd vendor I'm like yeah you can see my data I don't care. Your insurance companies are not going to say to the 3 pd we don't want to see in our data if we can help it right, but one way trust opens up another interesting door which is some of the hardest problems are around fuzzy things, even trans notion of a geofence is what does that mean, that long and go down like what 810 digits right, how exact where exactly is the center of your house that those are vague things in two way zero trust on a fuzzy notion like where exactly is your house on a geocode is tricky one way trust gets really easy. If you're willing to send me that geofence whack I can figure it out. So it's interesting to that scale to ask even if this is designed for zero trust. If we could in fact have a one way trust you can also scale up in another direction there too so also something to think about. Yeah, to add on to that. James we're looking at another proof of concept right now in collaboration with Mississippi in which they've contracted with a third party to go in an image. All of the affected properties after a catastrophic storm is hit, and the state has access to that information. And they're happy to make it available to insurance carriers so that's kind of the one way thing you can ask for it and you can get it. But the state has said but you know what would be really great is if we give you those images could you just tell us if that house was insured. That's huge. But even if you could tell me for how much. Now I can very quickly create a report for FEMA so they know what kind of assistance we're going to ask for. And again it's very simple we don't need a huge data set to make this happen it's a very few pieces of information. We just agree to how we're going to exchange it now all of a sudden the insurance carrier has the images the state has what they need to help their citizens and have what they need to go apply for FEMA grants. So again, keep it simple, and it just a lot, a lot of innovation happens when you keep it simple and you start doing these things iteratively. Absolutely. And we did get a one question in about kind of borders and I know, Joan and Ryan we've talked about this before but in terms of, you know, open ideal, and our community right now being, you know, just focus on the US but you know what, what do we all think this looks like you know expanding internationally, eventually. That's why we're with the Linux Foundation. Exactly. The most successful largest open source group in the world and you enable that the development of technologies in an open collaborative way. And so that's how it becomes international but Ryan's the one that works for an international company that it can actually put a solution on top of that. Yeah, absolutely. And, you know, we do with the few minutes that we do have left. And I think that's a great question that leads into, you know, we need more participants here and if anyone has thoughts about, you know, closing thoughts, how, how we can get more participants where, you know, what, what we can do as a community to bring more people on. So let me answer Roland's question since he did ask it. The only my quick observation just reading in real time is, it doesn't sound like you need a whole lot of carriers to agree in order to get your first win. So some things you need network effects you need to get a pile of people before you can demonstrate value, I would say find one insurer because if you already have all that stuff in your blockchain, and you're giving me off of them observer access, you just need one carrier to get their act together. So I would just say, just put maximum effort the one carrier to find I'm just guessing. Because if it grab it, if they're one insurance will grab and they can demonstrate value just like Jonah's pointing out that the two wins that we have on this thing so far in North here, they'll go to in Mississippi. It's like if you just get those so just an observation. I don't know how to do that convincingly because hey if they say no they say no but you only one went to really get the ball rolling would be my observation. So I would, I was doing the same thing James region role was a question and I think that's a great answer because certainly there are those within any organization we we've touched on it a couple different times on this call that basically they're there for responses to resist data collaboration in any way that seems to be, you know, you know, their default operation, but working with one and I think James what you're explaining about even just one one, you know, single, I forget what you called it, but the PowerPoint collaboration, I think is good. Don't let good be the enemy of great. And then go from there clearly there will be the more scale that you bring against a given problem is great. I would also suggest from a global perspective for your question is that use cases around the world, even if it's this one to one do have a reference effect right hey if it worked in North Dakota can it work in another state can it work in the UK. These type of addressing problems so. Anyway, I share your pain rolling to be honest because you know trying to get attention on some specific use cases and getting people to take action is the name of the game but you know start simple I think is one of the things don't reference here. Awesome. Well I want to thank first and foremost on all of the attendees today for joining us and our questions and if questions are rolling in here I will follow up with you for sure. And I can touch base with the panelists and provide answers for you. And I want to thank all of the panelists and it's been wonderful working with you and I hope we can continue to expand our community. Thank you. Thank you all so much. Thank you everyone for hosting. Thank you. I want to thank all of our panelists for their time today and thank you everyone for joining us just a reminder this recording will be on the Linux Foundation's YouTube page later today. We hope you join us for future webinars have a wonderful day.