 Good morning, everyone, and welcome to the August edition of the Hyperledger Financial Markets Mortgage Subgroup Meeting. Before we get started, I'd like to express my appreciation to the Financial Markets Special Interest Group and the Hyperledger Foundation for the ongoing support and just making this group possible. So definitely appreciative of all of their help and support. And with that, why don't we just dive right into the next slide. Okay. As always, we want to start out and note that this meeting is being recorded and is under the umbrella of the Hyperledger Foundation. So we ask that everyone abide by two items, the antitrust policy and the code of conduct. The antitrust policy states that we avoid discussions of company-specific products and projects. We don't make negative remarks about other companies and products. And for the code of conduct means that we treat each other with respect, never discriminate, and communicate constructively. We fully support Hyperledger's policy of openness, equity, and inclusion. And for new participants, we welcome you. If you'd like, just please introduce yourself in the chat. If there are anything, any specific areas of interest, please let us know. And we'll try and cover those either in this meeting or future meetings. So welcome, everyone. Okay. Here's our agenda for today. We've gone through the welcome meeting, excuse me, welcome in meeting housekeeping. Just a brief Hyperledger community announcement next. Then James will walk through the state of the blockchain in the global and U.S. markets. We're going to have the Reniri team and IBM team walk through blockchain and mortgage servicing. And then we'll go through future agenda topics. So I always like to go through the slide because this slide is just to reinforce with this whole meeting and this whole group is about. We're all in the same blockchain journey, although we may be at different points in that. This group is meant to help everyone on their blockchain journey to, number one, demonstrate the feasibility of blockchain technology through mortgage industry use cases and to define potential implementation paths for the mortgage industry. That I'm going to get off my soapbox, but I'll hop on at any other time. Okay. Community information. I just wanted to do a brief public service announcement. You guys have all seen that Capital Market SIG changes name to the financial market SIG. This update is ongoing. If you're getting multiple emails, multiple notices, please delete yourself from the capital markets mailing list. And I'm sure you're already on the financial markets mailing list. So let's just make sure that we get through this transition as smoothly and quickly as possible. Okay. Additional information that we try to cover in every meeting. The next three slides I always mentioned for people that are new to the group. This slide shows the different sites for the Linux Foundation, Hyperledger Foundation, Financial Markets and the Hyperledger Group as well. So these are great links for additional information, especially the GitHub. The GitHub is fantastic for Hyperledger. To access that information, you're going to need an LFID. So this slide helps you go through it. And the video is very helpful as well. I won't go through that in detail. Then the blockchain training, free training. I mentioned this every time I've taken this training. It's fantastic. Couldn't spell blockchain before. Now I think I can come close. Okay. That gets us to the state of blockchain in the global mortgage industry. And I'm going to hand it over to my good friend, James Henry. Marvin, thank you very much. I love your whole being able to spell blockchain. All right. So jumping into some of the new items in the mortgage industry and blockchain. Marvin, let's go ahead and move on to the first slide for global. So again, just a refresher for everybody. These are all articles that we've covered in our previous meetings. Last month's theme, we were talking about consortiums. You're going to see a little bit of a flavor of that in this month's announcements as well, in that there's a variety of companies that are starting to team up together on different initiatives. We've also, from the global perspective, we've kind of been traveling around from continent to continent on the planet. And this month happens to bring us back to Europe. Marvin, next slide. So starting out in London, Bank of New York, Mellon, and Goldman Sachs have successfully completed the industry's first agency securities lending transaction using HQLA distributed ledger platform. So the total size was in hundreds of millions of US dollars and HQLA actually created an ISIN level securities tracker called a digital collateral records or DCRs. From loan securities, if loan securities it received from Bank of New York, Mellon, and then Goldman Sachs provided a digital copy of those trades. The ISIN level DCRs are the very first of their kind in the industry. They represent specific ISIN quantities held in custody, and those records will enable holders and agencies to transfer ownership of any security on the HQLA distributed ledger without having the need for conventional settlement mechanisms that we're used to. Also coming out of Europe in Bulgaria, AQRU has announced a partnership with Blimp Homes to launch a crypto mortgage platform titled Bybrics. Blimp Homes is a home search and transition collaboration platform that rewards its users for browsing homes and it connects borrower, sellers, and real estate professionals. AQRU is focused on the DeFi space and they're aiming to provide a more efficient and secure way of providing financial services. And they believe there's a sufficient market for crypto back homes, noting that there's around 147,000 wallets out there currently holding $300,000 in Bitcoin or more. And AQR and Blimp both believe this represents a huge opportunity to manage crypto back mortgages. So we're going to be keeping our eyes on this one to see how it progresses throughout the year and we'll bring additional updates in future meetings. Marvel, let's go ahead and move on into the U.S. space. So, you know, similar to before, we've had a variety of companies that we have talked about over this last year of meetings. Actually, both of the articles that I pulled in for this month will touch on some of those previous companies. Next slide, Marvin. So there are three mortgage companies that are partnering up with Figure. You've all heard of Figure by now. We've talked about them quite a bit in these presentations. Guaranteed Rate, Homebridge, Synergy One, they've all partnered with Figure to leverage Figure's HELOC technology to create a new branded product. Figure's HELOC application process is entirely digital and the application takes as few as five minutes to complete and approval can come within a matter of minutes. Figure saw record demand for its branded HELOC product in the month of June surpassing over $325 million in funding volume. And the total was the eighth consecutive record month of Figure's HELOCs and represents a nearly 300% increase over the past quarter. Their HELOC volume was at 58% in Q2 and is continued to expected to climb in Q3 as demand for the product continues to grow. And we've talked about that in some of our previous sessions, that as we're watching what's going on in the interest rate market, that HELOC is a prime area of business that companies are going to be focusing on. Likewise, for Milo, if you'll recall, we actually first introduced this Miami-based fintech company back in our February presentation. They were talking at the time about launching a crypto mortgage product, which they effectively did in April of this year. And since that time, Milo has closed over $10 million in cryptocurrency mortgages with this product. Just as a reminder on what this product is, through the crypto lending program, Milo allows borrowers to pledge their cryptocurrency through regulated custodians, such as Coinbase or Gemini. They require no down payment other than pledging the cryptocurrency as collateral. And the interest rates right now are running about 6.95%. But those can change on an annual basis on this product. What happens is annually, Milo actually reviews the asset ratio of crypto assets that are pledged as a security to the loan amount. And that in turn directly has an impact on what their interest rate is for the following 12 months. They started out with roughly about 20 employees at the beginning of the year in this space. And they've already doubled that to 40 employees. And Milo has plans to soon introduce a crypto mortgage refinancing product. So again, keep your eyes and ears open for that. And as we receive future announcements, we'll continue to be sharing with this group. Marvin onto the next slide. Yeah. And here's just a reminder about the Hyperledger Wiki site. In fact, I'm going to drop into the chat the link for everybody to quickly be able to access it. I just want to do a reminder about this site. As Marvin walked through the links in the beginning, you can find access to all of those links from the site. You can sign up for the notifications. That way you'll get notified of our future monthly meetings, as well as when we're publishing articles up to the site. Also on the left hand side, you'll see copies of all our previous presentations. All of these presentations are recorded. So at any time you like, feel free to come to the site, go back into any of our previous months and see what we were talking about or discussing. Marvin, at this point, I'll go ahead and pass it back over to you. Thanks, James. That's some great information as usual. And now what I'd like to do is introduce our next speaker, Karen Belaza, the Chief Operations Officer of Renierry Solutions. Karen has an extensive background in mortgage and mortgage servicing, which includes being a founding employee and past president of Celine Finance and leading the company from a startup over 10 billion in servicing. Karen was also a managing director for CapMark Finance, formerly GMAC Commercial Mortgage, where she was responsible for offshore activities, including European and Indian operations and leadership roles for the company's REO disposition group for closure and bankruptcy units. She has an extensive background, so I'm going to stop there and just hand it over to her or we're just going to keep going. So I'm going to hand over control to the Renierry team. Welcome, Karen and the Renierry team. Thanks, Marvin. Thanks for that intro. I've chosen to forget a lot of what you just mentioned. Thank you. So I have here a group of folks from Renierry. Some of us will talk and others are responsible for a lot of the work that we've done so far, so I wanted them to be here for our little party. First, I wanted to just explain, if you don't already know, I think most people know who Louis Renierry is. He's the guy we named the company after and pretty well known in this industry in housing and mortgage. He's been long known for innovation in this space. As you may know, he's the guy who first conceptualized the securitization process, which really changed this industry, gave folks access to mortgages across the U.S. In some pockets, you really couldn't get access unless you happened to have a bank in town, so this sort of smoothed out the access to mortgage finance. For us, for this endeavor here, he brought together a team of industry veterans. I'm one of them and challenged us to come up with a solution that would transform housing. Of course, some of us, you just read my bio. I spent a lot of my career in servicing. I did have a few years in originations and also about 10 years in commercial, but in servicing. So we have a good view of where the pain points are in this industry. We spent some time thinking about the end-to-end mortgage process and where are the problems. And of course, it would have been easy to start at the beginning. That was my first instinct, was we'll start where the mortgages originated and work our way down the line. But to be honest, what we thought about and your vision statement for this group kind of got my attention too, because you talked about closing as efficiently as an auto loan and provide a superior customer service. We thought about that, but then where does the loan go after that? If you originated on blockchain, for example, where's it going to go? It's going to wind up in a not a blockchain system, but a mainframe system. There are two that are prevalent in the industry today. So we started with servicing. And as you go back one for a second, Will, go back to that slide. So the light blue boxes there kind of depict some major activities in servicing. And it is significant in size. And we talked to Marvin the other day and he made a comment. You picked the most complicated place to start. And we did that on purpose because we thought if you're going to leverage blockchain technology to do this, then the loan needs a place to go where it's going to live and the data provenance is secured and can be locked down and trusted. If you originate on blockchain and then dump it into one of those mainframe systems, you've kind of lost all of that advantage. So we started with building a servicing system. And as you can see here from the darker blue boxes, there are a lot of entities involved in servicing the loans. At the bottom of this slide, we've shown just a few of them, but there are multiple vendors. And there's the borrower and attorneys, regulators and insurance companies and taxpayers. So this is a pretty big system. And we're going to give you a look at it in a minute. You know, there have been a lot of improvements in the industry on in the servicing area, but most of them are around the edges. There's been new tools bolted on over time, but the core systems really haven't changed. And it's really difficult to change because of the breadth and depth of the system. And adding these new tools onto the sides of it create their own problems and creates problems in maintaining and managing and being able to make changes. Regulators have been putting more and more pressure on servicers over time. The ability to accommodate a new change or a new data requirement is really important. And for a servicer to be able to prove what they've done to have that information to be trusted is really important. So our goal is to create a more modern platform for housing beginning with servicing, leveraging blockchain to improve the homeowner experience and servicers economics. So it's a big a big charge. We are at a very cool place along your timeline, Marvin, that you were showing. We're almost at production. So we are we are pretty far long. We've stayed very quiet. We haven't been out on the radar. We are not doing press releases and talking about what we will do. We're going to get out and talk about what we have done so that you can be sure that what we talk about is coming to market. So I talked about why we started with servicing. This slide here just kind of gives you our view of blockchain and the value proposition as it applies to servicing. And I think, you know, this is this is kind of our guiding, our guiding light as we think about how we're going to deploy blockchain here. We talked to, we've got our service, our subject matter experts, many of them are here on the call today. We've talked to many servicers of all sizes to get some feedback on, you know, what what is the need? What would they want? You know, what would it take to get them to convert? We haven't been encouraged at every stop along the way to come up with something. They want to see an alternative. So our main objectives and this is answering the call from the servicing community. They want a cost effective system with easy navigation. They want the customer experience to be improved. And we're going to show you how that'll happen here in our system in a minute. They want to be able to accommodate regular regulator changes and requirements more easily. Today, if a new requirement comes out, an example one of one came out a couple weeks ago, you have to be able to accommodate a borrower's language preference, which means in the system, you need to know what the language preference would be. That is a change that while it sounds pretty simple, it's it's a challenge in some of the mainframe systems today. We can make that change very easily. In fact, it's already there. And we need to be able to maintain an immutable audit trail so we can prove compliance with the regulator's requirements. So we're going to show it to you now. We're we've got an end to end system that we've built. We're just going to give you a little glimpse of it because we don't have a lot of time here today. But this is the result of all that work thinking about what it is that and we're going to get Christy to to show you the demo. But we've gotten we had two primary servicers that worked with us just to make sure we're up to date on all the requirements and what needs to be done. We also consulted with vendors who are in that ecosystem, who who work with the servicers and are suffering an equal amount of pain from the systems that are out there today. So we're going to bring bring you this is the system life. So this is a this is the landing page, we call it. This is where you would start out if you're familiar with any of the other systems, you're starting out on a green screen. A few of them, they've got a new front end, they put a new front page on it. So you're not kind of looking at a green screen. But this is our landing page. And our goal here was to make sure that the person who's on this page who brings this loan up, maybe you're talking to a borrower, all the relevant information that you would need to do the job, answer the question and not make a mistake is here. So each of the items that are on the banner were thought through. And this will give you a quick view of the loan status. Any alerts or flags or warnings that would be on this loan come right to this front page. So you don't have to dig back. Now, the way it works today in servicing is you have to dig back into the comments and notes to figure out what's going on. So everything is here. If the borrower is called in recently and asked for something, somebody's working on something on the loan, you can see that. You've got all the balances and property addresses and anything that's going on on this loan would be right here. Christie's pointing at a series of modules that you can jump to. So again, in the other systems, you have to know a screen name in order to type that in and bring yourself to a default screen or the investor screens. They're all easily accessible just by pointing your mouse at the proper module. We also know from our experience that history of the loan is super important. So for any one of these modules that has some loans history connected to it, you can get easily to the loan history. You can get to a history of just cash transactions or just default transactions, just escrow transactions. All of that was built in to the system. One of the things that I love to show is documents. So in other in servicing shops today, you have a separate document management system. And this will be important for you to take note of as we show you the POC on the loan transfer documents live in this system and accessible from the loan and even from a module. So if you're in a foreclosure module, Christie, I'm sorry, I know I'm jumping all over the place on you. But in the foreclosure module, you can see foreclosure documents without having to leave the system and log on somewhere else and see if it's there. You can get a quick view of whether there's a document out there and you can access it from here. So that eliminates extra cost for a separate system and eliminates a lot of time in having to skip back and forth and find what you're looking for. One thing we'll show you real quickly is, Christie, you wanted to show, yeah, go ahead, property. Why don't you take over from here? Can you open your mic and talk for a second? I can take over, yep. So as she was talking about foreclosure and documents, so here I'm going to add a couple of inspection things for the property insurance claim that's open on this guy. It's as simple as this. And now you can see those two are here. And if I go over to property insurance claims on this open claim, go down to the bottom, it's now telling me that there are three. Let me move this bar out of the way really quick. There are my three documents right there. And then the other really nice thing is there are some things that you can order out of the system, inspections, valuations, and title orders. And it's as simple as I hit a button, I hit order inspection, I pick my vendor, my order date, my type. I can put a comment if I want to and I hit order inspection and then there will be a file that goes out and goes to that particular vendor that they can then receive into their system and then do the inspection and then send back that file and it matches up on that and that PO number that you see there and it will get the data back and update that order for the inspection. So you can see that just really quickly. Overall, I really like the UI. It's very clean. I like things simple because I'm so simple minded. So it definitely kudos on that. Every time I think about a loan servicing system, I think about it's just a huge rules engine. So tell me how you guys handle the rules and build up from there. And then I'm sure that there are going to be questions on what's on chain, what's off chain. I mean a whole mess of technical stuff like that. But we're coming into that in just the next slide or two. So yeah, the rules. Go ahead, Christy, take the rules engine, please. So the rules engine, there's a whole series of configs. There's roughly 40 of them and they each touch on the different sections that you see down in here. So for escrow, there's a lot of rules around escrow. For example, you have to have your cushions. You got to have your whether or not you release funds on an escrow overage. And some people will be more conservative than what the actual rule is. And so we give them the ability to even within their own shops manage their portfolio the way that they see fit within the confines of the rules. So without going to all the individual ones, you can see here, we have a couple of configs already built into the system, which allows them to do those rules. And then we have some more looser rules that are Merlin specific, which includes things like being able to put holds onto the loan to keep certain things from happening in the event of bankruptcy or foreclosure, right? And they may have a series of things that they want. They get to decide what those are. And then the system reads that and says, oh, this loan's gone into bankruptcy. And now no regular payments can come in. And, you know, whatever it is that they decide is the sky's the limit, really. And then we do have some other rules based onto the fields themselves, like you can't put alpha in a numeric field, those simple rules. That's, that's most of what the rules engine is is doing right now. Okay, great. Thank you. So I think that's all the time we have. This demo takes us about an hour and a half or so. So we tried to cut it down just to give you a taste for what it is. But I think Will's going to go into some more technical detail around the platform. Right. I wanted to be cognizant of the amount of technical material we have. Not that it's that great for today, but I think we're going to dive into some specific blockchain topics that might pique the interest of the group here. So this is just a high level view. And I'm going to ask one of our colleagues from IBM consulting to consulting to join in and Krishna can jump in and provide some kind of like the color commentary here as we go. This is a kind of a high level look at our target network. You know, starting out with the first version of the system, the idea is to get the services needs met. But while we're doing that, we're building a system that will provide future capabilities. And, you know, so I guess the question, why do we choose hyperledger and not trying to do a sales pitch or anything. I just wanted to, you know, go through the kind of things that we thought about and what made sense to us and what will be part and parcel of our application. You know, primarily, obviously the blockchain is providing tamper proof evidence of the mortgage record over its long life. And I think there's some folks on the call here who could take that and run with it in terms of what extra value that will provide. You know, in addition to that, you know, the mortgage servicer, the borrower, the vendors, regulators, they're not on this chart, but they're certainly part of this. They're all part of a kind of a de facto consortium of the way the industry works today. And so there are future use cases that can be decentralized, but for now our focus is on the tamper proof evidence. And you'll see why I'm leaving transparency out for now, but we'll cover a little bit later in the next demo where we're showing you a proof of concept of how we can transfer mortgages from one servicer to another taking advantage of the blockchain technology. You know, we saw some real-world usage in large-scale enterprise applications. I'll give an example. TradeLens is one of them demonstrating high performance and reliability under load and usage. I think that Hyperledger is well positioned for financial services. Some of the capabilities that we're using now and we'll use in the future channels and, of course, maybe private data collections. We have a way around that today, so that's not so important. It's open. It's modular. I think many of you know this sort of thing. Open chain code language is available. Plugable consensus mechanism. Raft, for example, with its election law replication, it provides crash follow-up tolerance, which is more than sufficient for the Merlin's usage. We're a private permission blockchain approach. Interoperability for future capability. Ability to handle ERP 20 tokens. Things like that. Certainly the publicly available documentation chain code pattern system. Wealth of information. It's available from the community and then, of course, the ongoing support. So, some of the reasons why HLF was our chosen direction forward. Christian can explain some of the technical aspects of the diagram here. Yeah. Thanks, Christian. Yeah. We trusted the nodes. You know, I wouldn't touch with the nodes. Usually within a consortium based blockchain network, a member organization maybe choose to participate directly or indirectly, right? If the member organization is a direct participant in the network, meaning that it owns or operates in order on the network and they have a sticker and may influence the overall governance as well, but they will have to consider the onboarding and operation costs. But if the member organization is an indirect participant in the consortium network through a platform or an organization here in the instance, it's Merlin platform. They will have lower participation costs in the network. And here multi-tenant is an indirect participation. So services like medium or large may choose to join or own their own node to participate. Or on the other side, they may choose to participate multi-tenant similarly for investors and vendors and not limited to these participants and but many more to join the network. So with respect to the context and a node here in the case of having both application components along with the fabric, hyper-logic fabric components. And with respect to the channel in the middle is the hyper-logic fabric which we'll just mentioned about its capabilities which can actually participate in the node set. Question, Christian. So if I interpret this diagram correctly, the investor node for example would be a node for investors. It wouldn't be a node per investor. It would be a many to one relationship. Same thing for all of the other nodes. Is that correct? Yes. So again, a investor node can be a multi-tenant node or a large to medium or medium to large investor wants to have their direct participation as I mentioned, can also own a node on the network. So that comes with the operational and onboarding related costs as I just said. I think the value here with Hyperledger is the ability to have for example a multi-tenant node where there's logical separation is a huge benefit compared to some of the other choices we had just in trying to operate a business and trying to provide some rational approach to a business model. There's no need for physical nodes and that's a big benefit. Yeah, absolutely because one of the things we ran across in an earlier POC is if we did have one company per node or one entity per node, then the cost of the blockchain just skyrocketed. So it's a much more efficient approach. Sure thing, latency performance becomes a major issue as you expand it out and there's no need for that. So that's good. But so let's take a real quick at some architecture just to give you a feel for what we're trying to do. We didn't want to get into too much detail on that because it takes away from the overall presentation. But basically we're following the platform architecture with IBM Cloud for financial services reference provides a really great base force because the industry requires at a minimum SOC2 type 2 and we're doing some other things as well. But the nice thing about the cloud environment is that the key services are I'll say out of the box quote unquote with the appropriate configurations are already proofed and are already part of the IBM SOC2 type 2 is very important for our client base. And as well there's some near HA service resiliency obviously that's something that has to be carefully dealt with in terms of putting the operational system together. But you know overall security by design is just baked in. So that's really a big help in getting the product launched as quickly as we can and being able to operate it. And then of course the application that sits inside of all that you know we like this they were approaching web 3.0 but we won't claim that we're there. But you know it's private permissioned blockchain open source software microservices IBM has been a great help with helping us pull that together. API interoperability is the core of the system. Everything that you can do every screen every interaction can be done through the appropriate APIs. Renier is looking at you know opening APIs where it makes sense with the vendor community things like that so that we can you know provide great support but also make it easy for the industry to do what it has to do. So all those screens that Christie was showing you different things everything is of course API based if you will. At cloud native horizontal vertical scalability Christian is going to talk about some of the details of the data platform. Code maintainability Karen talked about that before that's very important and of course it's partly the tools in the product we're using but it's partly the approach that you take. You've got to be able to make changes in the system quickly and that's all there. And of course you know from the application standpoint security and privacy by design and of course we plan to suck to type to recognize that this is an absolute. So I just wanted to touch on that really quickly without getting into too much detail. Will just a quick question on the open source software tools. Can you just speak really quickly on what your tech stack looks like. I think you kind of interspersed a little bit about that but I don't know if you can talk about that. Yeah I can. Well you know let's jump here first and then we'll come to the first of it. All right we'll start at this point. Christian if you wanted to explain this and then we'll move it up from there. Sure. Definitely. So part of this target state we got certain data store components if you see from top row consists of like two rows again. They know hyperledge fabric consists of two related data stores. One being the ledger itself. It's a transaction log keep adding the mutable related records and keep growing on that side of the ledger and the other component in the hyperledge fabric itself is the world state. It can be a couch DB or a level DB but couch DB helps enhance us in case if we want to have some kind of rich query but there are options there and so the world state will always have a current value whatever we are trying to store but in this case we are trying to store the proof of the transaction or the loan transaction and that particular component would take it off to store the current values and if we move down to the next row often consists of multiple stores again one being the Postgres SQL. The reason for that is like a lot of relational database activities we do see on the mortgage industry. So for that reason we are going to store majority of the data into the Postgres and that particular loan content will be hashed out and put on the on chain so that we have a proof going on there along with the off chain store on the Postgres SQL and if we see another component on below the Postgres is the object storage where it will actually capture all the documents like huge documents like PDF, JPEG, audio or any kind of media file and instead of putting them on chain or off chain which was to put it on some kind of an object store so that the large files can exist there and the hashes will still be correlated back into off chain Postgres SQL as well as on the on chain so that you know we have it referenced and similarly the in-memory cache predominantly for APIs any kind of lookup data for the UI related lookup data frequently used beta we plan to use that as well. So to add on to that you notice that's the data platform and I'm just going to drag this across real quick and just to well I think I'm going to do then you know this is a high level reference architecture that kind of gives you a you know Kristen just covered the data platform here and we're not going to get into the details here we don't have enough time but you can kind of see that this is essentially the system if you will so and it's you can you can get details of this just from a reference architecture standpoint you know it's not magic secret sauce I think what we're doing in some of these cases are pretty cool but you know it's a UI code there's application services that support it security underlines the underscores the entire system and we will be adding a data analytics module to this as well and we've got some pretty neat ideas on how that's going to be fed in a near real-time way so it'll be essentially real-time data without affecting the performance of the system so we think that's pretty unique and so this gives you kind of feel for it you know of course it all sits in sits it doesn't sit on top it sits inside of the the cloud so that's just a quick quick glance at that I will this is Angel just wanted to jump in here we got a couple questions in the chat here okay now where are the nodes being hosted on that's one question number two is are you using ibp or ibm cloud or just plain fabric I'm gonna I can do that yeah Christian why don't you take it yeah yeah sure well you know you can add you know in case if I miss anything so the nodes will be hosted on ibm cloud and as Bill mentioned earlier in the slide you know financial services cloud on ibm cloud again which takes care of the you know security and whatnot other other aspects of it so to answer your question the nodes will be on ibm cloud and in case you are looking on what tech component it's like a Kubernetes or OpenShift kind of component where the nodes will be getting on it the whole hybrid fabric will be installed on any of those so if our Kubernetes are OpenShift sorry in case I missed any other point what was the other questions for him I'm trying to yeah thank you for answering those questions there is one more question when a lender transfers to a servicer are you connecting to their LOS or importing or is it being done via a file dump well that's a version one versus a version two and beyond question so the answer of course to get the product out the door given the state of the industry today would be the file dump because that's how servicers work today but we have every plan in leveraging the API nature of the system to be able to interconnect I do we do have a system that's on our first it's first on the list but we're not going to get into that here today that we would want to have an API connection we have to look as the industry evolves as to how that would work obviously you're going to want to have an automated approach to that when the origination system is blockchain based but you know really the world that we're looking at as we get to the end of this year beginning of next year would you know the file dump is the way to go to get the product launched I hope I answered that I mean yes we have great plans on how to use this technology to interoperate if you will with with a origination system as well without the existing file dump process okay great I appreciate that and Paul bow and I you did you were the first guy I was the first question out there I just want you to know we're not leaving you hanging here if you can take a look at Paul's question and then we can answer it towards the end of the call that'd be great but the questions are coming in fast and furious so a clarifying question so you are not using IBM blockchain platform we are the IBM blockchain tool set is a critical part of the overall system and hyper ledger of course is the underlying so I mean it's important and Christian can help me with the distinction yeah yeah sure you know so not not you know tooling as such from the IBM blockchain platform where it helps you see IBP console has been open-sourced to hyper ledger fabric community right you know so we what we are going to do in this instance is we're going to take the additional tool set what we had from the IBP but on top you know taking the hyper ledger fabric by itself from the you know IBM has certified images we're going to take that comes with the tooling as well and we are going to install hyper ledger fabric support related images on top of the Kubernetes or open shift that that is where you know we get tooling as well as the open you know the open source of the hyper ledger fabric yeah we haven't made our mind up on the difference in the Kubernetes versus OpenShift that's a two great options yeah technically it's you know to the user to the system it's not a difference obviously to us it is but yeah all right tonight to have that there's a question about MISMO underlying data standard we've but we were you know to say we're using MISMO it would not be an accurate thing however we've taken the MISMO standard data dictionary into play Christy and her team have spent a lot of time making sure we've at least captured the data elements I'm not going to tell you that we're completely aligned with naming standards and things like that but as far as a the data elements and requirements are concerned we're doing our best to make sure all of that is taken into account you know it's a it's a huge undertaking to say that you are MISMO standard compliant and I'm not sure anybody could ever say that so but I don't want to leave you with the impression we're not doing it we're certainly paying attention to make sure that we meet those data requirements as they apply to our system so let's in the interest of time we want to get ready to show you a prototype that Reneary and IBM developed to show you one of those techniques one of those things that we think you're going to be really cool with the system and really provide some value and it's just a design it's called Merlin to Merlin loan transfer obviously you've got to have two two services who were on Merlin in order for it to work but it demonstrates some of the hyper ledger capability and some of the other things that we're doing and Karen you were going to explain why do we pick loan transfer so I think we put a lot of little notes here on this slide but I do want to say something that's very recent and and relevant for anyone on this call who is interested in originations and customer retention and you know all all the things that that are important to this industry you know there's a JD powers service customer satisfaction survey related to us mortgages that was released about a week or two ago yeah he's got it up on the screen and two of the three noted issues that are appearing in the survey results and I think they had over eight eight or nine thousand borrowers who responded to it their problems are around loan transfers and in fact if you dig into the numbers in this report you'll see that customer satisfaction drops about 20 percent when a loan is transferred and that's because it's a difficult activity for borrowers in the way the loans transfer from one servicer to another you know I don't have deep your knowledge of servicing but even if you transfer between two servicers who are both on the same system it's still like a brand new thing coming over because the data is used differently by everyone even if you're on the same system so the data mapping and the the loss of data and miss misunderstanding of the data as it comes across is very critical borrowers it's this survey says only about 15 of the borrowers whose loans are transferred will think about going back to the original lender you imagine that you did all that work to originate a loan you get him on the system get him to the servicer and then you he loses confidence in that original lender because of this transfer process so this is critical and we want to show you how we see in the future when multiple servicers are on this platform how that transfer will happen and I think I think you'll understand then why we think even though this transfer process doesn't seem like a big deal it is affecting the the borrower confidence in the servicer and the original lender so thanks Karen so here we have a kind of a just a quick kind of but not a deep dive but a dive into the underlying data platform that we were explaining and we can talk this through Christian did you want to just walk through yeah details here sure uh yeah I know the idea for this is to prove the architecture that it can scale for large volumes we can intend to keep a proof on blockchain so auditors and regulators can validate the authenticity of the data at any given point and ease of loan transfer by enabling both services and to get to see their respective data without any delay and with appropriate access restrictions at at each level of confidence like you know we have web api and spot contract layers for each of those components each application node have the application stack components like apis web and UI related and connected to an off-chain store like as I mentioned Postgres and SQL Postgres SQL and object storage related components and in middle if you see the blockchain node consists of the Hyperledge factory components such as peers, smart contract, pledge yards and certificate authority and auditors many more part of that the factory component in a framework so you'll see key key items here we're hashing the proof of the data and what's even the actual data in the Postgres database just to boil it down to the essential and so with that I think it's time to have the IBM team pull up the demo that we built let me stop sharing here and we're going to show you the Merlin to Merlin loan transfer prototype which will eventually become part of the production system but we wanted to do some testing and things like that so it's a limited scale down thing to really take a look at the Hyperledger and Postgres interactive interaction and to gauge the performance and make sure that things are going to work as advertised and that the channel technology does what it's supposed to do yeah thanks well I'll take over from here thanks Spencer yeah this is Spencer from IBM so I'm going to be demoing a transfer of 100 loans you know from a servicer a to a servicer b so let me go ahead and share my screen and walk through this yeah and while he's pulling it up just remember this is just a limited prototype that doesn't have full functionality it wasn't designed for that but I think you'll get the point as he walks through how it's going to work so prior to opening up the application what I've done is on board the loan data of the 100 loans so what we're looking at here is the representative data of 100 loans so you can see in column a we have you know from 100 down to 199 is the IDs making up the 100 loans they're all going to be tied to servicer a in the system and then you'll see here we have you know 50 or so representative fields about that loan now you know in a real scenario we might have you know 1000 or even more fields but you know we've boiled it down to 50 for the sake of the demo here so now that the loans have been boarded into the system I'll go ahead and open it up and show you what that loan transfer process looks like so I'm going to log in as Alice who is a representative from servicer a that's funny we have an Alice in our test system as well so Alice will see here a table of the previous transfers but I'm going to go past that for now and I'm going to create a new transfer so you'll see here it's asking me to input various data about that transfer so I'll go ahead and do that here and we're going to transfer this today now you know in the real world this might be staged for for 15 or more days to you know meet regulatory requirements you know to to alert the customers that their loans are being transferred but for this it'll be you know a same day transfer so we're transferring 100 loans from a to be and then I'm going to input a little bit of information about who's buying those lines so the investor so lastly I'm going to upload the loan numbers that I'd like to transfer so what I've done is take a spreadsheet here and you can see that I have the 100 loans and just the loan numbers from servicer a so we're not inputting all of the data here it's it's just the numbers so I'm going to upload this spreadsheet and and again in the future maybe it's not a spreadsheet maybe we're selecting the loans we like to transfer from a UI it's just an early prototype so I'm going to initiate the loan transfer so now when I hit initiate all of the data related to these loans including I didn't I never mentioned I've also uploaded two documents for each of those 100 loans you know again there might be 10 20 or more documents in a real scenario here each one has two so when I initiate this loan transfer all of the data will now be visible to service or be including the documents and all of the data so there's no physical transfer of the data from a to be it's what it's a permission switch on the blockchain when I hit initiate here so I've gone ahead and initiated that and let's go ahead and look at it so this is our transfer you can see you know again all the data from a to b it's pending and here we can see a table view of the 100 loans now this is highlighting some of the the important fields in this table view we can click into them we can see the documents that I've uploaded you know we can download these documents view the documents in this case we have a balloon mortgage note and then you'll see here we have the old loan numbers which are related to service or a and then we have the new loan numbers this is blank this will be filled in by server survey so what I'm going to do next is log out of Alice from a and log into Bob from b and show what the acceptance of this transfer is going to look like this ability to collaborate and confirm between these two servicers before it actually goes is is going to resolve a lot of the pitfalls that these borrowers are falling into during the transfer process and and also dramatic reduction in effort from both sides yeah one question from a hash perspective so when is the hash going to be created for that transfer and I assume it's going to be at the loan level not at the servicer level that's right it is at the loan level and once it is accepted then the hash will get generated and yeah and get stored onto the chain okay so at the loan level okay and so when the new loan number is created and added to this page for example that's when the hash is created and went to the block uh no um so that is the mapping done uh but you know there is an actual action taken on the screen I think the Spencer is going to do that um click the ax up then the service of these actually accepting the transfer then the hash of the loan will be generated and get stored okay understood thank you and so you know now Bob from B you know he's seeing all of the data um that was so Bob is going to review all of these loans uh you know and make sure that that these are the hundred that he would like to accept in the transfer so he can view some of the the primary data in this stable format here you know or what he can do now is go ahead and export this to CSV and here he'll have a view of all of the data available for review now what he's going to do after review is tag servicer B's loan numbers uh to to each one so I'll go ahead and do that and we'll just call this um the old number plus servicer B and again while Spencer's doing that just to make the case this is not how the real system will work we just took some shortcuts make it easy from a prototyping standpoint to use the spreadsheets for upload and download however a lot of servicers work that way today so that's not to say we wouldn't have that feature but we certainly have an online capability for this sort of thing so Bob is then going to upload um you know his respective loan numbers um and submit those loan numbers and and now you'll see that here they're they're in the system um with with each of the new loan numbers so the the loan transfer isn't complete yet uh you know B will need to accept the transfer um and in this case when he when he clicks accept it will be you know a final immediate transfer of ownership from A to B so I'll go ahead and do that here and so at that point uh servicer A will still have a record they still have all the records that they created over the period of time they were servicing but they won't have any of the new activity created by servicer B because they've picked up responsibility for it so B gets all the historical information uh and plus is the active servicer going forward A still has their record so after clicking complete you'll see now the transfer status is complete we see that the previous servicer was A but at this point now the ownership and the future history of the loans will be tied to servicer B so the last thing I'll show here is just I want to go back to Alice's screen and show what what they'll see you know as you know Karen mentioned A will see all of the the previous history of that loan for very you know regulatory or reporting uh compliance requirements but anything that happens to the loan past the transfer will only be visible now to servicer B so and A sees a zero balance so now yeah you'll see the the UPB here for A is showing zero as it's now tied to servicer B it feels like a dream transfers are are so bad in every way for the servicer and the borrower and the investor you know we've had Ginny May and CFPB and all of the all of the GSEs they've all come out with a lot more scrutiny over the servicing process and it's because of of the borrower issues that get created in the way it's done today so maybe this is Karen for you or Christy there's a question on the chat is there a cash settlement feature that calls funds due to or from current to new servicers for advances escrow balances etc yeah good Christy so there will be a fund settlement but uh if if you're if the question is related to is Merlin going to do that specifically I think uh for this first go around the expectation would be how it's done today so there'll be reports that corporate accounting can get the cash will transfer like it normally does and then at some point the future when we get to that on our roadmap yeah our favorite thing is on our roadmap for sure so I think that we just you know Spencer did a great job of just going through that and showing you step by step and obviously underneath that as Christian explained the blockchain is doing its thing and of course there's some programming some API calls things like that smart contracts of course that are doing the writing of the hash to the chain and but with a you know a little bit of development work you have a significant capability and it was just again we just use it to highlight the consortium the channels the hyperledger the the approach that we're taking and it shows you some of the future capabilities that you could extrapolate from that and so that's you know if there's any other questions anything like that I'm happy to help answer and I think I'm going to share the next page here which has some contact information on it and with that I'll just throw it back over to Marvin and thank you thank you guys that that was an excellent presentation one thing that really stood out for me as we were going through those slides in the demo is you guys made a concerted effort to make architectural decisions that address responsiveness from my personal experience the servicing platforms that are out there unbelievably slow and I think you guys went to great pains to address that so definitely congratulations on that that that looked fantastic well thank you thank you our approach you know we didn't say much about postgres because the highlight today was hyperledger but certainly with the scalability and the parallelism on hyperledger and couple that with some of the I'll call it pseudo vertical scale horizontal scalability with replication and things like that that are available in postgres today I think we're going to have and of course you know following the the architecture as we've described I think we're going to have a system that works very well under a heavy load with you know millions of loans and thousands and thousands of users and hundreds of thousands of borrowers logging in through the portal however that goes yep so thanks for that if anyone else has any questions for the Ranieri team I'm going to go ahead and leave this last slide on there you guys can reach out to them directly with any questions or comments if you want you can also just post questions on the discord channel for the hyperledger mortgage subgroup and we'll get those answered if they're general questions around this but overall thank you Karen will Spencer and Krishna this was a fantastic demo definitely appreciated thank you everybody this this was a great presentation I've actually been getting pinged with a lot of comments as well directly to me very impressive what you guys are lining out I appreciate you guys coming and doing the demo and Paul and follow up to your ltv question on Milo wasn't able to find the answer while we're on this call but I'll continue to take a look and see what I can get back to you yeah that's great um I will um because I actually I wanted to follow up with my law on some things anyway so it would be good for me to go in and not be dumb when I first call them I appreciate it no problem hey uh the last thing I wanted to cover because we are already past our time is the next meeting is September 8th that's going to be James Showning and Casey Rock from the Hyperledger IM project they're going to demonstrate their Hyperledger wallet and then the next couple meetings we're tentatively talking about property records and smart contracts and then a blockchain panel discussion potentially in November maybe December so um with that uh I think we've gone through questions I just want to say thank you to everyone apologize for running long but I think it was definitely worthwhile thank you everyone have a great day thank you all thank you take care everyone have a great day everybody