 Okay. Welcome everybody to this Asia Time Capital Market Special Interest Group from Hyperledger. We have an amazing speaker today. I'm very happy to introduce Tessata who has been, I think I've seen in the blockchain community since way back and has been building this company. He is the CEO in Tain and I'm going to hand it over to him. Just before I do, just two quick things. As in this is a Linux Foundation meeting, I have to remind everyone that we do have an antitrust policy and we also have a code of conduct. Please do look at those on our website. Pretty common sense though. Please treat everyone with respect and with that I'm going to hand it over to you, Tessata. Hey, thanks Julie. Good to be again part of a Hyperledger session. I will spend maybe 20-25 minutes on the presentation but in between also if there are any questions, happy to take them. Otherwise, let's just spend some time discussing at the end because I don't think the presentation is going to last that long. I don't think you all expect that to last that long either. So let's just first look at the problem which we are attempting to solve. So the structured finance industry A is too big and too critical and the metrics of that metric of it being too big and too critical are one of the proof points of it being too big and too critical is that it was big enough and critical enough that it could pull the world economy down in 2007. So the first thing to look at is how big and how critical the structured finance industry is to the world economy. Second, how non-transparent and how inefficient it is that there are about half a dozen financial institutions which interact with each other in every structured finance transaction and all of them interact via Excel sheets and emails. So it is a very big and very critical industry where everybody is interacting via Excel sheets and emails and the result is that an investor has a clear view into his or her portfolio about six weeks past the event. So for example around 15th of June an investor will get a report about May which means that if something happened on first of May the investor is getting a picture of that occurrence or something that happened to the assets underlying that security on 15th of June so about six weeks later. So it's non-transparent and inefficient. And then the question is that why this could not be solved for so long. Now one easy way of solving any such problem is that if I can have one database across all these parties which is the lender, servicer, trustee, issuer, investor, rating agency then I can automate any function. Everybody feeds off the same database and it would work what in IT industry we call like a SaaS solution. But the fact that even 15 years after the global financial crisis institutions have not accepted any SaaS solution means that they are not willing to run their transactions on a central database. Now if people are going to keep their own systems but we still want to ensure one version of truth, obvious option is to do it on a blockchain. So then if you see it's a big and critical industry, it's non-transparent and inefficient and the only way we could solve this problem was by bringing this industry onto a blockchain which then ensures that there is one version of truth across all these parties and then we can write smart contracts to automate all the interactions between them. So that is really the crux of the problem that in pain is trying to solve. Now this in a way summarizes each of the problems that we face which is siloed data because everybody is working of their own systems, layered costs because every transaction has to be reconciled and you keep adding more and more layers of opacity and with each layer of opacity you are adding a layer of cost. All these processes are manual and we talked about transparency. Now couple of points about how we have approached solving this problem. First is that it's very fashionable in our industry to say we'll create a consortium. Now you look at all consortiums across all use cases in blockchain that have been created. Hardly anyone has got above 8-10% of the market share and you can look at trade finance or any other industry segment where blockchain has been used, there's been an attempt to create a consortium. So one approach that we took was that how do I have the same end goal as a consortium without actually having to form a consortium which means that having to go to 10 different kinds of entities and convincing them that become a part of consortium. Now good thing for us in structured finance is that the structured finance industry significantly consolidated with the trustees. For people who have understanding of fund admin or fund management industry, it's very similar to how the fund admin fund management industry is consolidated with custodians. So it could be BNY, Mellon State Street, each of them could be around 30 trillion dollars instead of JP Morgan, Northern Trust. So within first four or five of the custodians, you would have covered 30-35% of the market. The consolidation in structured finance with trustees is greater. So any partnership with a trustee is like a quasi consortium. So that is the first approach we took that will partner with a trustee, treat it like a hub and like a quasi consortium and take the solution to the market. The success of this approach is that now we are working with the number four trustee and the number seven trustee and we are in conversations with top four. And as you can see on the table here, the number four trustee has 10% of the market share, which means that my starting point itself becomes that I am running a blockchain on which 10% of the industry will run because the trustee is my partner, they are the hub and all the other institutions are the spoke for these transactions to run on a blockchain. So that's first the conceptual model of how we run our blockchain platform. Couple of points about technology. I am not an expert but our broad approach has been to solve an industry problem, not be a blockchain solution, which means that I decide that, okay, these are seven different entities I have to bring on to the network. Then I start looking at each entity's pain point. So if the custodian or the verification agent has to come on to my blockchain so that the original loan pool that has to be securitized is verified. The best option for me is to give him an option of completely automating the due diligence on a loan pool. Now that requires artificial intelligence, computer vision, NLP to be able to read loan contracts and hence the solution is not blockchain solution is AI. So you provide an AI solution for the investor. Investor is used to really state of our data visualization and analytics tools. So you provide the investor that. So it's basically looking at every entity and not going to that entity and saying that please come on board our platform because we are a blockchain. It is to tell them that we have a better software solution than whatever you are used to and yes, then the additional benefit is that we are eliminating all reconciliation. Everything is immutable, etc., etc. because of the blockchain. So that is one big difference we had in terms of the approach and we have taken this approach and lot of things. For example, onboarding. To onboard data, yes, it's good that everybody is a node but if they are not a node, we have provided for Excel sheets to be uploaded. We have provided for API integrations for data to come in. On the output side, yes, it's good that if the investor is a node but if the investor is not a node and they are habituated to asking their admin to print a PDF of a report, we have provided a PDF creator for that report online. So it's really small, small things which ensures that any resistance to adoption is eliminated. That you give me an excuse why you can't come on to the platform and I have a solution for you. And in summary, what it has done is that it has looked at four, five of the key problems of the industry and we have solved that if you look at them even as a point solution. So the due diligence and loan documents by a verification agent, I talked about standardization of the service at loan tape, which is a big issue where we looked across the word. The best practice was the ESMA regulations in Europe. We made it as part of the standard platform. So we solved for the loan tape standardization problem, deal modeling and payment waterfall. We do using smart contracts. That's another problem we solved. And then I talked about data visualization and reporting. So if you just ignore that there is a blockchain underneath as a software solution, it is solving some of the biggest problems faced by the structured finance industry. I'll just spend 30 seconds each on some of the points that I've talked about. So this is the solution using artificial intelligence to actually read a loan pool. So it's basically using AI to automate due diligence in a structured finance transaction. And that is eliminating almost 90% of the effort which verification agents were putting in. And what it does then is that when I'm onboarding a loan pool onto the blockchain, it's like what the blockchain industry would call an Oracle. I am getting verified data onto my platform from actual documents from which the data is sourced. So I'm using AI to create almost my own Oracle to get data onto the blockchain. Then this is the biggest challenge and the biggest differentiator for us. Structured finance industry is called structured finance because the deal structures are amongst the most complex that you will ever get. We take from the actual indenture, which is the contract for a structured finance transaction and model every single term as a smart contract. So generally these structured finance pools are divided into multiple tranches. Stop and the safest tranches may be invested into by pension funds and insurance companies. Bottom tranches may go to hedge funds, etc. Each of them have their own terms, their different trigger events. All of this is coded as a smart contract. Now what we have done is that if we had said that we are blockchain only solution, many of you who have kind of more technical insights than I have would realize that blockchain is not a great analytics platform. So while we take the input data of all the collections around loans coming in and we standardize them, we run this row on the blockchain, which means using smart contracts. But here we pull the data from the blockchain into an analytics data store and then run a standard data visualization and analytics tool on top. So then what the investor is getting is A plus B, which means that how much payment that needs to be made to each investor and how trigger events and all deal terms are done, they are being calculated using smart contracts and hence they are immutable, auditable, etc. But that is still not constraining the investor from doing on-demand analytics, which otherwise would not be possible just with smart contracts. So investor can, and since it's like traditional data visualization tools, create any slice and dice of data because we have pulled the data out of the blockchain into a data store. They couldn't have done it if it was still in the blockchain. This automation of the models. And then there are other entities who were consuming this data by sourcing data from let's say the trustee or the servicer like the rating agency. Now the rating agency just has to login and it has verified data available in real time. So for those of you who have seen the Big Shot movie, which is about the problems of the industry, you can see that this is how it solves each of the problems which was highlighted by that movie, including the data that the rating agency is getting because now rating agency is seeing exactly the same picture that the issuer is seeing that the investor is seeing or anybody else is seeing in the industry. So in summary, these are all the parties which are being brought onto a common fabric-based platform, which because of the nature of blockchain, there's nothing unique that we are doing which ensures there is one version of truth. Then there is complete process automation, some which is kind of routine software process automation, the other through smart contracts. This transparency and compliance explain to you that the whole deal terms is automatically getting calculated using smart contracts. And this platform which we call Intane Admin, Intane Admin will continue to run on hyperledger fabric. We are creating a replica of Intane Admin in Avalanche to create a tokenized layer on top of that which we call Intane Markets. And as I said that from a time when you did a $200 million transaction on a blockchain platform and you went and did a press release about it, we have brought it to a stage where almost every month we are onboarding two or three transactions onto a platform, each of them would be anywhere between $100 to $300 million. So this is slightly older version of the presentation. So we are close to about $5 billion in assets under admin on the platform about 17, 18 transactions. And this is what I was mentioning that there are across industry use cases, there are very few blockchain platforms which are getting more out of a smart contract than we are getting. And simply the reason is structured finance industry has the most complex structures and we are modeling the whole contract with smart contracts. So all the terms that you see for each of these deals and these are actual deals running on our platform, all these calculations for each of these deals are being done and executed through smart contracts. Now, as I said that yes, intane markets won't run on fabric, but it is important to keep in mind that any tokenization actually depends a lot on what you are administering underneath. So for four years, three years, all of us have been talking about security tokens and there is hardly any mass scale adoption of security tokens. And the reason is, and I've taken the mortgage back security example here. So this is how a mortgage back security is created. What any tokenization platform does is just takes this last step. So everything is done. There is a token ready to be created. You create a token and then offer it to investors. Now, blockchain is pretty useless. If all that you're doing with the blockchain is to create a token at this last step because the problem is not at this last step. Yes, tokenization ensures liquidity, but the precursor to liquidity is trust and transparency and tokenization in itself. So we do not know any more about a token on blockchain than we would know about that instrument of a blockchain just because it's a token. And hence, the bigger problem is here, but nobody attempts to solve it for two reasons. One, it is not as remunerative because you get trading volumes on tokenizations. And second, this is not cross-leverageable. So when I say cross-leverageable, it means that you can create a token platform where you're issuing real estate tokens, structured finance tokens, any kind of tokens because all that you care is create tokens and put it on a platform. The admin is very, very asset class specific. So even an asset-backed security to mortgage-backed security, it's not that easy to replicate. And hence, it's a very complex structure for administration of each asset class that you have to create. And that is why we have spent almost four years, 20 to 25 developers at one point in time solving this problem in structured finance that get this part of the industry onto a blockchain. The last part will automatically take care of itself. And as I've shown with the numbers on assets under admin, number of kinds of deals, et cetera, that we have on the platform. It does prove that that's a successful approach. And next month, we'll be launching in Tain markets. So yeah, that's about it. This is the last slide to show what our vision is. As I said, that deals under $100 million are unviable. It takes many, many weeks just to get a transaction going, get a real picture of the transaction. We believe that we will be able to reduce the deal size or viable deal size to a 10th. The fact that some of the intermediaries in the industry are willing to share 50%, 50% of their fees with us, which means that at least 70, 80% of the transaction has to get automated for somebody to say that 50% of their fees they'll give to us. So these are already proof points. And this is really the end goal in this. So yeah, so that's kind of the presentation. Happy to take any questions. Excellent. Excellent. Thank you. And I think Biffin is here now, right? Biffin, do you want to say hi? Yeah, thank you. Sorry, I was a couple of minutes late, but I'm here now. No problem. Anyway, great presentation. You know, the heavy lifting that is associated with the process automation, which is the main point that you're making in your, and the success that you're having. But and you mentioned one thing in the end, which is that a lot of these are bespoke. That means you've created something that is specific to structured finance. But there are elements of it which are common across, I mean, you showed the mortgage platform, for example. You know, there is a prospectus in mortgage, which is very similar to what you sort of automate in terms of reading the document with AI and then taking the deal structure and then automating it in smart contracts. I mean, my question would be, isn't there a lot of such commonality across, like even in syndicated loans or some other kind of, you know, these kind of structured finance product like mortgage bonds. Isn't there a lot of commonality there? I know that, you know. Yeah, sorry. Yeah, I know that you've said that they seem different. But in a sense, there is a pattern there. There's things that you could do commonly. So you could take your product into other kinds of products. It's what I feel. But maybe you can tell me why not. Yeah, so syndicated loans you talked about. So for example, in CLOs, most of what you saw is mostly CDOs. So in CLOs, the role is of a collateral agent who holds the collateral. So at the conceptual level, yes, there are, I think in US, it's called LSTA. There are LSTA structured documents of each of these loans which need to be read and those are collateral which need to be maintained. So at conceptual level, it will look very similar. But when you're structuring or designing a blockchain, there is a completely new role that of a collateral agent in that transaction. So you will not be able to run on a common blockchain platform, or you will end up creating at least half or dozen, I guess in Hyperledger fabric, it's called Org's, which would be relevant only for a specific type of transaction. So there would be a lot more complexity that would come in. And even then you would at best explore some adjacencies to credit derivatives, but not beyond. So in implementation, it gets very, very complex if you're trying to administer more asset classes. And that's not only for structured finance. If you see traditionally, and I ran a JV with State Street as a CEO, in State Street or Bank of New York, which are biggest fund administrators in the world, how they administer fixed income funds is completely different from how they administer equities funds, the software systems are different, the teams are different, everything is different. So at the administration level, it's very, very difficult to create commonalities across asset classes, whether blockchain or otherwise. Okay, I'll buy that answer. Obviously, some of the logic and some of the smart contracts could be taken from there and built used with other bits to build other platforms. Is that right or not? Yeah. And one thing is that, yes, it pushes the envelope in terms of what you can do with smart contracts. So when we started or even till second, third year, when we used to struggle, we used to talk to folks at IBM, and they would advise us against using smart contracts for these complex deal structures they set in fabric or anywhere smart contracts were never created for you to take a 200 page contract with so many terms and convert everything into a smart contract. So it's what you're doing is not even advisable leave alone us helping you out for that. But at least we have proven over a period of time that you can do that. And to that extent, it's replicable in other areas where people may have been a bit different about using smart contracts to automate far more complex structures problems. I said that a couple of questions and before that, thanks for a wonderful presentation. It was very insightful. So the first point I was just trying to understand the current solution which you're talking about. So if I'm correct, is that from your solution, I am one of the counterparty between for the structure finance as a client. I may not necessarily be dealing with the counterparty in anything which is related to in a tokenized form. So I will continue the way I'm doing the business. It may be on Excel or in the document which I'm creating it. Once the deal is done, that is something which is getting tokenized or you are also enabling the deal itself which initiates which can this platform can be used for it. So that's why it's a kind of two layered platform. So what we run on fabric is Intane Admin which is administration of a structured finance transaction. So the transaction actually happens the way it currently happens which is off-chain. Once the transaction happens, the life cycle of each of these deals is 7, 10, 15 years. So that process runs on-chain which is the whole servicing and admin. So again, if I refer to global financial crisis, so what you saw in global financial crisis is not the problem when the deals were happening. The problem is that I have invested into a security. That security is based on a tranche. That tranche is a tranche of a 10,000 loan pool and that is dependent on somebody's ability in Florida to pay for their mortgage. So I have to realize that value of my security is linked to somebody's ability in Florida to pay for their mortgage and how I completely make this link transparent is what the whole servicing and admin automation. The tokenization layer is on top of that which is a separate layer. So in that case, you are able to do this, facilitate this post transaction because you have a one common entity which is your trustee, just like a custodian, which basically it binds both the sides of the transaction just like a custodian. So the follow-up question which I had, sorry if you were saying something. No, no, finish it. So if I'm trying to compare it to something which is a typical OTC derivatives market for that example, is that a good similar case study or case use case which can have a similar platform which can be built around that but as what I understood from you because every deal may be done through a different counterparty and the SSI basically settlement instruction would also be different between these two counterparties. So there is no common as such a body which basically brings them and something which can give me a scale of economy for bringing this into a more a DLT platform kind of thing. Yeah, so the use of trustees is less because there is a common party to which is interacting with everybody. For example, trustees don't really interact with rating agencies. Trustees is more in terms of our GTMR go-to-market that the trustee is allowing me to access 10% of the market in one go. So for example, if I were to shut my shop today means I reduced in 10 to a two employee company in 10 will still continue onboarding deals because that trustee is 10% market share and that trustee will keep selling more and those deals will keep getting onboarded and once the deals are onboarded then all parties who are involved in that transaction will keep getting onboarded on my platform. So this schematic that you saw about trustees being hub is less of how a transaction happens. It's more about our go-to market that we have just used them. As a target, yeah, basically your market. As a market access point, yeah. Yes, that's correct. So in that, sorry, just maybe the last question if I just want to give a chance to others. So in that case, who becomes your kind of who contacts you to become a member of that platform? Is it the client themselves or how does it work? Yeah. So as of now, because as I said, trustees are part of every deal. So the deals would come to the trustee and if the trustee has moved to our platform, then deals would by default move to our platform. So we are not doing any selling for individual deals. The deals are being onboarded to our platform by default as long as we can get trustees to realize that it's in their interest to move to our platform. That was your question on. Yeah, absolutely. That's right. Yeah. Thanks very much. Yeah, thank you for that great presentation and I know the journey of you from day one. Yeah, congratulations for that success. Thanks, Kamlesh. Good to have you on the car. Can I interject for a second or is there other people who want to ask questions? I think go ahead and people raise your hands and turn your cameras on speak. Please just get involved. We have a bit of nothing. Yes, please. Yeah. So you're basically saying how the go-to-market and deals get created. Now the second part of it would be how a secondary market evolves from this and also the performance of the deal which you obviously becomes important at that point. It's like the mortgage in Florida, the guy is able to pay every month or was he late or she late on the mortgage payment that becomes important for the secondary market, a secondary market. I mean, thousands of such deals aggregated together in the case of mortgages but in this case these are huge loans that are made to specific companies or is a collection of loans that become loan obligations. So could you speak a little bit about the secondary market and how the transparency helps in pricing these instruments in a secondary market? Yeah. So we for ourselves also have been doing a lot of experiments on this. Obviously, everybody who's trading a token in secondary market is not going to do the kind of drill down that we provide to an institutional investor in a routine course. Obviously, they will have access to this level of analytics. So what we have experimented with is that for each of these deals create almost like a NYSE stock ticker kind of thing which is aggregating data and creating an indicator. Now, so that's just at conceptual and software level. As we launch and contain markets, we do not intend to make the tokens to be tradable. So they will be yield generating tokens at least for next 9 to 12 months because we want to be absolutely sure about regulations before we offer them into the secondary market. So that's why I wanted to divide into two parts. At conceptual and software level, we have experimented with aggregating this, creating machine learning based models so that even if the investor who's trading in these tokens in the secondary market is not going to be spending 10, 15 minutes drilling down all the data that we provide them, they still have an NYSE ticker kind of thing about each pool. But the real implementation of this is at least a year away because we are not going to allow our tokens to be traded for now. Coming back to that point, so unless you track the performance, that means not just the issuance of the loan or CLO or whatever, but the performance. That means the payments that are made by the debtor and the tracking of not only the payments, but also the rating of that company itself. Sometimes that affects the trading in the secondary market. So are all those things also taken care of in your absolutely. So I'll show maybe I may have some screen. Otherwise, as we are talking, I'll pull out to show. So the investor can drill down into every single loan and every single repayment of every loan. And in case of loans which are based on collaterals, get into every single property of each of these collaterals. For example, and these things that we don't intuitively realize is that we'll say that I want to figure out my exposure to a particular state. Now exposure to a particular state has two components. I may be living in New York, but my property may be in New Jersey in which I've invested in and hence bought in some other state. I have to figure out what is the implication of COVID on my pool. The implication of COVID is related to where the borrower is. Implication of property price dropping is related to where the property is. And the borrower and the property could be in two different states. Now these are things, nuances we have realized as we have started tracking the pools on our platform because that is the level of granularity to which we can track data now. So you can really, as I said, go into every single loan, every single repayment of every single loan and all the collaterals associated with that loan. In fact, we did a small pilot in India on a 5000 motorbike loan pool on which we even tested out things like whether the vehicle is hypothecated, not hypothecated, insured, not insured. Is there a correlation on the pool performance based on these properties? So that level of data is available. Lepin, I don't know if you can hear him. Yeah, I mean, I heard what he said and I'm trying to think about the level of granularity. So basically the last financial crisis happened because of liquidity crisis, which is based on people not knowing whether those mortgage securities really had mortgages that really were working and which tranche, but this solved that problem. So mitigate that problem when we have a very presumptuous claim to see that it will solve, but it will mitigate, means it will create far earlier warning signals. As I said, the easy thing is to think of the big shot movie and think of how much of what you see in that movies related to information asymmetry, right? Means everybody has their own versions of truth and everybody is doing their things and so only as a movie viewer, you realize that everybody is doing their own thing. Now that is something that the platform completely eliminates, that everybody knows exactly the same data. So that would create so that property market is crashing, people are not able to pay, that is not something that the platform can address. But you cannot have a situation that I as an investor realize that my pool became really defunct three months back or four months back when the Florida real estate market crashing, that won't happen, that will become real time. So I said, in addition to, as I said, other benefits that today in US, so we have 18 deals worth four and a half billion dollars roughly, that gives you an idea of the size of a deal, right? $250 million on an average. Because deals smaller than this are just prohibitive. If you are doing so much analysis, such complex structures, everybody reconciling manually, you cannot run a deal smaller than $100, $150 million. And we genuinely believe that in the next three months, we will be able to run a $10 million transaction on the platform. But the underlying are corporate loans, right? Or something else? I missed the beginning of your... No, no. Underlyings are in case of MBS, obviously, mod gauges. Within mod gauges, because US is a funny country where there are all kinds of structures with... So within mod gauges, they are fixed and flip loans, which have their own deal structures. They are HELOCs, which is home equity line of credits, which have their own structure. So we have multiple structures within residential mod gauges, commercial mod gauges. In asset backed securities, we have auto student loan, et cetera. Corporate loans would be much easier. They would generally go as CLOs. As I said, that it involves collateral agents. So those structures are slightly different. And the incremental value immediately till we create a collateral agent role on our platform may not be as great. So you're basically dealing with the originators who... We are dealing with the issuers. So originators in some cases may be the issuers. So we have very little to do genuinely with the originators, because the loans... So the lending is not happening on our platform. We are looking at creating eval so that the moment the loan is given, the data is captured on our platform. But till that happens, we do not have much to do with the originator, unless the originator is the issuer themselves, which means the originator has another legal entity which is packaging their loans and selling it. So for us, the starting point is the capital market cycle, which starts with the issuer rather than the lender. Yeah. So the originators often package up the loans into a bundle of some sort and then sell it to the... Whoever is going to do the issuance and the issuer will may buy from multiple such originators and then issue a bond based on those cash flows from all these different bundles. I mean, I worked in mortgages, so I'm familiar with this process, but it's fiendishly complex only because of the fact that the data from all these individual loans, which are only tracked on a monthly basis because the payments are monthly. And then by the time it hits the bundled loan, it's later, one month or something like that. So month plus one month later. And we used to run a lot of analytics based on the what each bond contains in terms of the vintage, where it originate from the, all kinds of MSAs, metropolitan statistical areas where the loan originated from all kinds of other data like that. So that level of granularity is normally very difficult to handle because there are so many different loans backing a single bond. Looks like you're dealing with that kind of granularity. Absolutely. In our case, any data that the servicer has or any data that the issuer has, which is about the collateral. So as I said, there's something that European regulators did really well. Early 2019, ESMA came up with their securitization regulation where they created Excel sheets where every cell is marked and described in the format in which data has to be reported. It's as bureaucratic as only European regulators can be, but we took that as a standard and made that the standard asset class wise template on our platform. So we are capturing data to that level of granularity in the US now. And where does this data reside? It resides on different VMs which form each of these nodes. Meaning they are on the blockchain or somewhere else? They are on the blockchain. It's predefined frequencies for the analytics part get pulled into a data store, which is what I explained, that you can't run a live analytics or on-demand analytics off a blockchain. It's just not feasible technology-wise. So for that purpose, the data keeps moving into analytics data store for the analytics to run, but what has to be run through smart contracts, which is the payment waterfall, etc., runs within the structure. So isn't that quite a bit of data to keep on the blockchain? What are the challenges you faced there? Keeping data on the blockchain is not that big an issue. As I said, running analytics off the blockchain will be an issue considering the amount of data that is there and that's why we don't do that. Okay, forget that analytics, but even to just store it, what do you use? You're using some private data collections or something else or is it all inside the fabric channels? It is inside the fabric channels. So each deal is a channel and it's inside the fabric channels. Okay. So typically, what is the volume of such data inside? It will be like hundreds of thousands. So each deal, some of these deals, maybe a few thousand loans, each loan may have up to 30 data points being reported every month. Some of the deals we have on our platform are two years old. So you can just do 5,000 into 30 into 24 months on a deal. And you said the duration of some of these are like 15 years. So if you... We'll go on for 15. In fact, we onboarded. So one of our trustees took over a deal from Wells Fargo, which is a 2005-06 deal. So yeah, so those are like 15-year-old deals that we took over. But I'm also talking about the deals that you currently administer that could go into the future for 10-15 years. So you're saying something like 30 data points monthly for 15 years. 30 data points for a few thousand loans in each deal monthly for 15 years. Yeah, that's a huge amount of data because I've always heard people saying that it's difficult to do that on fabric and to manage that kind of volume. I'm not even talking about querying an analytic part of which is what you said is taken off into a separate data store. Yeah, so this is just to give you an idea. So this is just the Excel template for each deal. So these are the number of fields for each, this RRE is residential real estate. Now all deals may not have all these data points. So this Excel is the European regulators Excel, which is what our template is based on. So all these deals, so till now here, it was only the loan level information, now collateral, which means the information about the house because the loan is different from the house. So you need the information about the loan, then you need the information about the house. So this is 110 data points. As I said, all deals don't have 110 data points, but potentially up to 110 data points in a single deal and residential real estate. And you have let's say 5000 of these. 5000 of these, yeah. And every month, every month. That's a whole lot of data. Yeah, as your bills are completely going haywire, so that's something we have to figure out what to do. So that's one more question. So in terms of when you have onboarded any of the deal and there is a payout say monthly or whatever the frequency has been set up at each payment, is that the payment triggers, what is the mechanism or what is the role of a platform of say there's a tranche which needs to be paid every month. So currently, yeah, so currently our platform does the calculation exactly how a trustee's internals or paying agents, internal software would have done based on which the payments will be made. The payments are executed manually. That's as far as Intain admin is concerned. In Intain markets, the payments will get executed automatically. But because of that, Intain markets will allow to begin with only standard deal structure. So the big difference is that currently the deal can be anything that the participants in a deal wanted to be. They do draft the contract. We take the contract and create smart contracts. In Intain markets, there will be standard sets of deal terms coded as smart contracts and the people underwriting a deal can only choose those. And that is important because in Intain admin, like anyway would happen on Excel sheets or whatever software systems, there is a gap between calculation and somebody paying. And you have a chance to catch an error, make a correction, etc. Because in Intain markets, the payment will get executed based on the smart contract calculation. We don't want to be in a situation where we are custom coding deals every month. That smart contract can't be put to test as rigorously as it ought to be because it's executing. So there it will be like a library of smart contracts, which you will have to choose from. Okay. So in that case, when you have an Intain market and that's payment, which you're talking about. So it basically triggers the payout for whatever frequency. Stable coins. So how, maybe if you take a minute to explain, I'm not very familiar with, say, if I'm, I have... If you're the investor, yeah. So if you're the... So we don't use stable coin for any purpose other than the fact that US dollar cannot move on a blockchain. So I have to find an alternative, which means that the moment an investor invests in a pool, that US dollar investment is converted into USDC. And the only stable coin we allow is USDC, which has two advantages. One, obviously, as everybody knows, backed by US dollar, dollar for dollar. Two, because we are doing this on Avalanche, it's natively minted on Avalanche and we don't have to use a bridge. So it automatically converts into USDC, obviously telling the investor that as you are executing this, I'm immediately converting to USDC. The moment it hits the issuer who has to collect that money, USDC converts back into USD again. Similar thing happens with the servicer. So servicer who's collecting the monthly payments, US dollars immediately gets converted into USDC. On USDC collections, the smart contract based calculation supply and hence that USDC is distributed to investors across the tranches as per whatever the algorithm has been coded into the smart contract. And the moment the investors receive their monthly collections money, it will immediately convert back into US dollars. So basically just for the period that it is running on the blockchain, it will be in USDC. The moment it hits the party, it will become a US dollar again. Okay. So when it actually comes to the bank for investors, it's no difference for him, right? If we expecting a USD flowing into his bank on a month of, fifth of month, it will be done in the same fashion. Absolutely. With obviously disclosure, currently USDC tracks to USD very, very closely. Possibly the transaction cost difference is less than a normal FX or a wire transfer difference, but we still in interest. But there's still a one-on-one kind of conversion ratio. Okay. So today, if this is on the in-tain market, which we're talking about, you're expecting that to be in a year time. No, in-tain markets will launch next month. But these tokens will be yield-generating tokens, which means that the investor invests and investor keeps getting monthly income as per the structure. But the investor cannot go and trade it in a secondary market, which is what Bipin is referring to. So what is delayed by a year is the ability to trade these tokens. The in-tain markets launch is due for next month. Okay. So as we speak, if somebody has onboarded and they are on your blockchain, so this is something basically making assurance that one of the deal, which is there, which is more of secured and which is basically single truth of source. But beyond that on a month-on-month maintenance, this is something which they're not relying on this purely on this data because there's all the transaction happening outside as well, which is not completely in the payouts, which is not integrated with this blockchain at this moment. So that's an understanding, right? No, no, no. They are completely relying on this data. It is not automatically executing the payment. The payment execution is off the blockchain. In case of in-tain admin, means I don't know if Fabric platform even allows to automatically execute the payment. I don't know how it would happen. So in-tain admin runs on Fabric. In case of Fabric, it is the single source of data for any payments that has to be made. But the actual making of payment is not done on the blockchain platform. In case of in-tain markets, which will run on top of in-tain admin, there the process of making that payment also will get executed automatically. So the software does not only the calculation that $3,500 has to be paid to Abhishek, the software will actually execute the payment. That is what will be different. Yeah, possibly the market data which the platform which is using, say for the calculation based on the library or any other benchmark, whereas the actual payout which is happening, it is how it will be very same or it will always be same because that's something at this point in time. The calculation, the payout is happening outside the platform. So that's something maybe I'm missing on that point because I understand when you have this integrated payment, USD token converting into USD, then it's a seamless thing, right? Whatever the calculation which the platform is doing, that exact amount which is being paid to the investor. But at this point, since it is happening outside the platform, you may have one value, say which the platform that you have a platform has calculated to one value which is showing in the cash flow, which may have some variation based on if there is a variable component in the payout or the market change. Basically the market data means any data means whosoever is going to make the payout wherever whatever the additional component is, that additional component will also have to be captured by the blockchain. So there is no calculation happening off blockchain to make any payouts of the blockchain. Just the manual task of executing a payment is happening off the blockchain. There is no parallel system running, no parallel calculation happening, no live or data input coming off blockchain. The number on the blockchain is the sacrosanct number based on which the payment will happen. Got it, got it. It's clear now. Yeah, I thought I was comparing it both to system running in parallel, yeah. Thanks so much, Satya. It spilled over the hour. It spilled over the hour. That's a good thing, right? It's a great thing. For me, it is not a good thing. It's midnight. Anyway, yeah, I think we have to thank Siddhartha for showing up and making this wonderful presentation. The other thing I wanted to actually say in the beginning of the call was that we are in the process of changing the name of the Special Interest Group from capital markets to something more in keeping with the kind of work and the kind of presentations, the kind of concerns that we have been airing. So we are going to probably call it, in a very generic way, the finance segment. And we are going to make changes because now it not only contains the long-term capital markets being the long-term markets. We are also focused on shorter-term markets. We are focused on D5. We are focused on financial infrastructure, which is what it seems to be. I mean, and most of us are concerned with that. So to group all that, to embrace it all, we are going to call it the finance segment. And I'll be circulating some kind of a proposal for that and to get the agreement of all the members. Thanks again, Julian, of course, does a wonderful job and I don't even need to be here. But this will make it all worthwhile. Thank you for your big effort, midnight. You're amazing to do that and it's great to hear that the SIG is expanding and there's so much work going on. So everyone should get on the mailing list, right, and get involved. There's so many activities. Yes. Yeah, thanks everybody. Thanks for inviting me. Have a good day. And Siddhartha, we hopefully will meet, you say, you come to New York often here. Yeah, July will catch up, look forward to that. Yes. Independence Day. No, second half of July. Yes. That's why when I'll be on the beach, all right, all right. Thanks, everybody. Have a good day. And thanks a lot. Good luck next month. Take care, everybody. Keep safe.