 will be shared with with the group. I hope some more people join. Looks like, go ahead. So roughly how many people can be joining this session? It is always unknown because this doesn't require registration. So we wait, you know, for like three minutes. There are people joining right now. We have already gone up to eight participants. Okay. Okay. So as soon as we cross, you know, 1003 or 1004, that is a New York time, then I will do the administrative trivia, which is basically talking about the antitrust policy and the code of conduct. Actually, let me do that right away. So all the calls in Hyperledger follow the antitrust policy by the Linux Foundation, which basically means that we are not here to collude to fix prices or to do anti-market activities. The other one is the code of conduct. The code of conduct really talks about the way in which we should conduct ourselves in such meetings. One thing is that we can disagree with people without being disagreeable. So that's one aspect. The other two aspects, there are many, but most of the other aspects are things like you are not going to quote someone without attributing nor hog the time and start talking about irrelevant topics. These are the only restrictions. Otherwise, the meeting is completely open and I'm thankful to get Contida and Teresa on the call. I hope I'm pronouncing your names correctly. I was reading the ADB paper, which consists of two parts. The one about Intanum that Tom was mentioning just a moment ago. The other is about the Scripless Bonds, which I suppose the presenters will go into greater detail on. It is a tough period right now, especially in the US, and holidays are coming up. So people are not flocking to these calls and they're always, we used to have something like 20 participants. We seem to have dropped quite a bit in the recent past. But no matter, we are going to have all of this material available in public. So the whole point of Hyperledger Capital Market SIG is to create a resource that would have both presentations and recordings that can be used asynchronously. Anyway, I'm not going to hold forth any further. I'm going to ask Contida and Teresa to start their presentation. You can share your screen and you can, you know, it's best to allow people to ask questions in the middle if possible. Otherwise, you run through your presentation and then we can talk about questions. Thanks again for showing up and making yourself available. Talk about this important topic. All right. Hello. You guys can hear me okay? Can you guys see the slides? Yes, we can hear you and we can see the slides. All right. Hello from Thailand. I'm Contita Aditya Trangason. I'm the project management officer of the DLT Scriptless Bond Project. And I'm also here with my colleague from the IT department, Chawisa. She oversees the IT part of this project. So as you may know, we have been doing the DLT Scriptless Bond Project and we have started our journey in, you know, as a POC. We built a pilot prototype system first back in the 2018. It took us about six months to finish the prototype system. And we proved that the concept is workable in practice and it can be implemented into production. So it took us from roughly about two years from early 2019 to 2020 for us to implement the DLT Bond platform, which is the platform that is used to reduce the bond and sell bond, government bonds. The project successfully went live and it supported the two issuance in the past, like in roughly about 25, 26 August. The new platform supported the sale of moving forward issuance and one bad bond issuance. The first on the moving forward issuance is it is the traditional way of issuing government-saving bonds for sale. We opened for sale from 26 August and finished the sale on September 11. So it was open for sale for individual and non-profit organizations and the value worth roughly about 55 billion, but it took us about only nine days for us to come to sell out the whole 45 billion, but worth of bonds to the public. The total transaction was roughly about 13,000 transactions. The value for each transaction is roughly about 1.8 million, but that's for individual and of course for the non-profit organization it is a much higher, like about 12 million, but so that's the traditional issuance. We also opened for one bad bond sale and that's the second issuance of one bad bond sale and we opened for sale roughly about the same period and that was sold out much sooner, but the value that we issued for sale was a lot lower. It was only five billion, but it was open for sale for Thai individuals aged 15 and older through the E-Wallet application that was developed by the Grung Thai Bank. The aim of this kind of bond issuance is to allow bond sales to reach a vast majority of Thai people, of Thai investors who are underserved financially and for investors to be able to purchase bonds through E-Wallet. That's the way that the Ministry of Finance they were looking to go, so that's why they implemented this service. So what's unique about this bond issuance is that the par value of the bond is set at 1 baht, but the minimum purchase is 100 baht, so that's even roughly about three dollars if you and the minimum amount that you can purchase bond is only three dollars roughly. So as you can see the average value per transaction for this kind of bond is far lower, it's only a 0.27 million baht per transaction. So this is the overview of the bond issuance that were issued through bond platform. I'll talk to you about the more about the business process and then the IT architecture afterwards. Okay, about the this project, I would like to talk to you first on the parties that are involved in it. Firstly, the BOT, the Bank of Thailand, we act as the platform and service provider of the BOT bond platform. Basically, we govern the platform and operate the system. We provide common services and we provide operational support for our platform and common services. Then we have the, as you can see on the right, the public debt management office who is under the Ministry of Finance. They are the issuer of the government bond in this platform. We have the Thailand security depository with the central security depository and we have the selling agent banks and that in this phase we have four banks who sell government-saving bonds that including the Bangkok Bank, Kroong Thai Bank, Kasi Korn Thai Bank and Sayang commercial bank who also owns the, actually each of us within the this circle owns separate nodes. So that role within this network is to establish nodes and connect to the platform. They have to regularly maintain operational health and they have to comply with security standard and operational guidelines that the BOT set. So basically, we develop the the system and we really develop the standard and we provide those standards for our members to develop their nodes within and connected to the platform that we built. So these are the main, the direct members of the DLTE bond platform but we also have the Thai BMA who is also responsible for assigning the Thai symbol code, Thai symbol which is the domestic code that is assigned to bonds. So this is the overview of the participants within the network but only Thai BMA doesn't own the node but they connect to this network via API. Okay, I've roughly talked to you about the role of each participant. The issue of the public definition of this, they have the role of to assign the bond information and selling criteria onto the bond platform. At this stage, the bond registration function within the platform is not yet went live so we still, we're currently working on it now. The issue also responsible for register bond information and selling criteria and announce bond for sale. The selling agent bank's that role is to sell bond to investors through their own channels whether they be a mobile or internet banking counters or a team. That role is also to collect money from the investors and transfer money that they receive from subscriptions to the BOT and then they have to provide delivery confirmation to investors. They also need to update the bookkeeping the records, customer records as well. In terms of the Bank of Thailand, we have the role within this network is we verify, sell transactions and allocate bonds according to subscriptions. We transfer money that we receive from the selling agents to the public debt management office and we also notify TSD who is the security depository to deposit bonds onto the investors account. On the TSD role, they have to assign the ICN and CFI code and they have to open security accounts for the investors and deposit bonds onto those accounts. Finally, the Thai BMA, the assigned bond, Thai symbol as I mentioned. I'll illustrate more clearly in later slides. Before we implemented the bond platform, we have thought about it quite a lot and we discussed with our stakeholders, we were working in groups using the design thinking method and so that we can not only build a new system but also discuss with them and re-engineer the process that can be agreed upon all parties that I mentioned earlier. So each of them have to make a little bit of adjustments or if not a lot of adjustment I specifically refer to the selling agent banks who have to change their process quite a lot. So this slide I would like to show you the previous process before we implemented the bond platform. This is the previous process that we have and you will recognize that it is very very slow and very inefficient. First of all, if investors want to purchase bonds, they have to send the bond subscription to selling agent banks and whether that be any channels that selling agent banks provide and then selling agent banks once they receive bond subscriptions from the investors in terms of money, they collect the money and then transfer the money to the BOT at the end of the day once a day. The BOT collects all the money so we act as the BOT act as the paying agent of the government bond so we collect all the money and transfer to the issuer within that same day. So the issuer would get the money from bond sale at the end of the day but however in terms of bond subscriptions to be recorded onto the security depositories or onto the bond book of investors, it can take up to about 15 days and that's because the transactions that are done by the investors are weighted to be submitted to the BOT on the weekly basis in batch file. So selling agent transfers the subscription to the BOT every Tuesday and then the BOT would collect all those subscriptions and then deposit the bond subscription onto the TSD every Friday. So at this stage it can take already up to T plus nine and once the investor's information is recorded onto the TSD, it can take another five days, six days for the selling agent banks to update the bond book of the investors. So totally can take up to T plus 15 for bonds to be deposited onto the investor's account and that's quite a long period of time. So we decided to re-engineer the process. We think about new process and we think about the new platform altogether. So this is how it looks like after we implemented the new process and platform. Customers, investors can still purchase bonds through the same channels that they previously had with the selling agent banks but instead but including the mobile banking counter ATM and internet banking. We also have the bond direct app which is the application that is provided by the public debt management office for investors to directly purchase bond directly onto the Ministry of Finance. And we also provide another channel, the Evalet channel that supports the sale of the one batch bond. So once investor purchase bonds through the selling agent bank channel, then the bank would send the subscription to the BLT bond platform that's within the big circle here. We have the Q management system. This is a non-blockchain system, but it's very important one since time sequence is very crucial in bond allocation. Bond has to be allocated on the first come, first serve basis. So we implemented the system so that each subscription is timestamped and then weighted onto the queue before it's allocated and be written on the blockchain. So once the bond is recorded on the blockchain then it is weighted until the end of the day, then the platform would notify the selling agent banks to transfer money to the Bank of Thailand and that's done via the botnet system which is the high value payment system that operates by the Bank of Thailand. So the selling agent banks would transfer money to Bank of Thailand and then at that day and then on the T plus one the platform would notify the TSD to open security accounts in case that investor do not have the security accounts or directly deposit the bonds onto the investor accounts if they already have their accounts. So the deficit process is finalized within T plus one. So once it is finalized then the selling agent banks can confirm to the investors on the on T plus two date. So as you can see the process can be shortened much by much from T plus 15 to T plus two. So the final process of bond allocation of the of our project ends on the deficit part but we have yet to support other transactions that are done in secondary market including the transfer of ownership or the coupon payment and stuff are still being done within the existing framework meaning that the selling agent banks or the custodian banks they have to send the inquiries to the TSD using the existing network the PTI system and transfer bonds over there and and those transactions are not recorded back onto the bond platform. So platform only acknowledge the the deposit tree of bonds in the primary market only. So this is our current process. Also I would like to talk to you a little bit more about the bond registration part which we have yet to go live our our systems. So this will be hopefully will be final we'll be finished by the end of first quarter next year. So in terms of bond registration the PDMO over here they would have to assign bond information as a link criteria onto the bond platform and then the bond platform would then send the notification to the Thai BMA to request for bond symbol. The Thai BMA then assign bond symbol and send it back to the bond platform for the BOT to notify TSD to assign ICN and CFI code and then TSD then assign ICN and CFI code accordingly. And then after that the once all the information is gathered from Thai BMA and TSD then the public debt management office can then publish bond for sale and after that the all the selling agent banks can then inquire bond information from the platform. So that's that's it for the bond registration. So that's that's the end of my part on the business part the business process. Any question from you guys? Money has a question on the chat which says why do we need a blockchain for this looks like the blockchain is only between key stakeholders not visible to selling agent or end users but it looks like it's visible to selling agents right or am I? Yes it is. All the participants within the network can can inquire for bond information and bond sales and they're only for the selling agent banks they don't see all the information but they only see the the information that on the need to know basis only their customer transactions or the public information like such as bond profile that they can see. The reason we we use the blockchain is that we we have it's more about capacity building at at the Bank of Thailand we have explore potential use cases that blockchain can be applied it. For instance we also we also exploit in internal project which is the cdc project and we also exploit it in this in this project as well and there were times that we considered many use cases and we we were actually not seeking for the right use cases but the the timing was right to establish a new system and process because it was so inefficient and we deemed that the the Scripps bond project is it's not a risky one it only focusing on the small group of participants only eight participants but the knowledge that we can get can be quite valuable for our for any few projects that we will undertake in the future. The benefit that we see from adopting blockchain is that from what we study is that like the it could enhance resiliency and transparency of the of the network for instance if one node fail then other nodes can still perform smoothly and moreover if data that can be in terms of transparency data has to be verified by more than one party within the network so transactions cannot be mended or forged so that that's what we learned before so that's why we it kind of make us want to study more about blockchain. It's not just theory you actually issued the bond and sold it to people yeah that is to put it in context 50 billion but that is about 1.5 billion dollars yes that's about right okay so 1.5 billion dollars has been has been sold through this platform now are we going to take a break here or are we going to go to the IT part how do you want to run this because there are people I'm sure who have questions so it depends on you and your colleague so and any more question from the business side otherwise I'll let my colleague learn through the IT part first and then you can ask any question you like again this is many again are the selling agents having do they take a node in that glt platform or they just do api they do all their inquiries they they have their own nodes and they can inquire where the glt bond platform directly okay thank you so you're probably using some kind of private data collections so that everything is not visible in hyperlogic fabric right anyway we'll go into that detail in when we talk to Chavisa okay I'll let you listen from my IT colleague first maybe you can get more idea on it yeah yeah and we you know there are a bunch of questions about other things too but is anybody else have any questions so let's let's proceed then and thank you okay let's hear from Chavisa then yes thank you anyway for I mean thank you for showing up and and explaining how you're going forward okay okay thank you thank you I'm Chavisa and today today I'm going to talk about the glt bond platform in the IT part as Queen Kanita said has already explained about the our participants as you can see so this slide is made quite a bit deeper so you can see that we have a blockchain network that composed of seven nodes of our part stakeholder and in order to connect to the blockchain so every every stakeholder will own the API server for connecting through the blockchain and as you have just already listened to the Queen Kanita so there are a few business data that can be illustrated as you can see from this slide so for each box is represent the role of the stakeholder so for the TSB for every star from the the left side but so for the BOT or as we add the list distribution so there will be some information like the bond information or the payment data or the sales transition that we will interact with the DOT bond platform obviously as you have just already listened that from the first form from the product part that there will be including with the ICIN or acquiring the CFI code so there will be so you can see that there are there are some data that we have to certificate first and then we have tried to compromise and then we separate the type of the blockchain channel into two parts the first one is we see that some data is quite the secret it's confidential right so like the sales transaction data we cannot it is it has to be confidential for each selling agent so that we separate we set the private channel for for for Chanel which we compose of the register registrar live DOT and PDMO and TSTM only one selling agent for each Chanel and the second type of the bond channel is the bond registration each channel you can see that all the stakeholders will be joined because the bond registration is will be the part that is related to the bond information so everyone will access the same information together and maybe i will jump too fast and actually let me let me introduce let me speak first and if you have any question you can ask at the end of my presentation okay so after after we have decided the channel as as i show you from the previous slide so this one is the social architecture that we use with all the all the stakeholder it means that each stakeholder will have the same social architecture like this so for for the common use case like the bond selling part this i will i will start from the bullet one first so if the user like the end user maybe as as you can see from the pilot slide maybe the people by the bond via the mobile channel right so the data will be sent to the new web app and then and then this new web app is mean like the the in-house system of the selling agent like bank of Bangkok Bank and then the data will send to via the API and it will send to the send to the part of the sales and reservation system so after after we has already do the the the quota allocation the one quota allocation the data will be will be involved to the to the blockchain and then the result of the quota allocation will be sent through the notif to the to the to the to the number five here and will send to the notification service after that the data we will collect in the event store and it will transfer to through the in-house application of the selling agent so this part is for the high throughput transaction of the bond selling as time is crucial and the technology that we use in this project is also as you can see from here so for the for the selling and reservation system part this part is include about the bond quota allocation so that we have to use some kind of rapid mq and radius to enhance the the the high high high throughput of transaction and we also use the blockchain hybridized fabric which also we also on the IBM blockchain platform for collecting the data of the sales transaction and also help to automate some kind of the open account transaction with TSD and also created the bond with the TSD as well uh okay so if you guys uh and and also um but actually this part is uh is about the the high throughput transaction but for the normal operation like we have to inquiry data from the blockchain okay so let me show this slide so uh the the end user or maybe the bank can also inquiry the data via the API as well here and so they can also invoke the data from the blockchain and receive the data back here or some actually there are also some type of the the business the business transaction as well that maybe the the selling agent also submit the sales transaction from to through the centralized system right but after they submit the transaction through the centralized system they also receive the result of the open account via the blockchain here so you can see that there also this line as well this this line is the event that we use uh in the in the in the blockchain for for for task for sending the data from the blockchain okay okay that's that's the end of the it part if you guys had any questions please feel free to ask so you said uh come let's just ask the question about the number of organizations i think it had according to what i understand there were seven seven organizations right yes seven yeah okay so uh so come this here actually so this all organization in the IBM blockchain platform or across different clouds yeah all of them on the IBM cloud IBM blockchain platform yes okay but but in the architecture there's one organization peer one peer two or there are different seven IBM blockchain platform means the seven different organizations right yes different organization but yeah yeah each of them will own their own account on the IBM blockchain platform okay okay so the most most interesting part about this is the way in which you guys are going forward by integrating aspects of existing platforms uh existing legacy system and only uh disrupting some parts of the system for example you're not automating the uh let's say oms of auto management system on the blockchain you're creating a separate um separate uh q the rabid mqq for the for that for the high value throughput and then your uh and the other interesting part is the way in which you have segmented your uh solution using multiple channels because that is a uh architectural feature of the the platform you selected which is IBM i mean which is IBM blockchain which is based on topology of fabric so these two are you know sort of artifacts of what and how you you know the platforms that you chose is it possible that you could have done this with private data collections or it would become much more complicated uh since the data is the data privacy is quite important right now right so actually we replicate some part of the the personal data into into this centralized system so in the blockchain we we didn't keep the private data i mean like my my own address or my own that's not what i'm asking i'm talking about the way you've designed the system which is basically having those channels uh the separate channel with each so there are let's say some privileged nodes which are basically the uh the bot the issuer and some other nodes like pdmo and the tsd and then for each for each uh bond agent or a selling agent you have a separate channel for the same information payment information and depository information uh and then uh you have a public channel which is basically public data about the bond the bond registration which is basically the ison the and the tie bond id and several other bond characteristics what the question i'm asking you can you do it with one channel if you had private data collections which are which are features in blocks in ibm blockchain or in in hypogea fabric that allows you to create those channel like structures using something else and less complicated channel setup i don't know i mean this is just a question about what's possible at at at the time that we we have implemented this project a private data has just come so it's quite a little bit new for us and actually and we actually um if we have to consolidate all of them together maybe maybe that there are some sometimes maybe it's possible yes but but but yeah i get it i get it at a particular point in time when this was still uh kind of new so you chose this approach uh normally that and and go ahead yeah and as and it's quite if you have to make sure that all the data will be segregated i mean like if we do the private data collection and we have to prove a lot that uh you ceiling agent will not see each other data i mean yeah but but for the pro actually for the for the for the for the for the share if you separate the channel it's quite clear cut yes for for presenting to to all of you as well true uh so it depends on the number of selling agents if you have many many many selling agents then this approach becomes very slightly more difficult because each new selling agent needs a new channel and then you know as as the platform scales with the number of selling agents then you need to create new channels every time so there's a there's a you know cost benefit to this approach and since most of the most of the sovereign bonds are only sold by some important banks then this approach is perfectly fine for the for the use case that you have right i mean if it is a more widespread sort of a bond that has to be sold by many different parties then then you got to have you know a different approach and things are changing so that's that's great uh anybody else has any questions otherwise i'll ask some more yeah i have one question is anybody there yeah i i i get it okay go ahead okay so this is the private channel so i think we can see the four entities and uh three looks same and fourth one is a there's a requirement from any some kind of party to have a separate channel like suppose bbl kbn ktb those are banks or some kind of sales is it yes selling agents right yes yes okay so but for example right now only we have a five channel and suppose number of channels increase maybe so for example maybe uh thousands of sales as n it could be or it could not be maybe in india or maybe in the bigger countries it could be the scenario but the sales then could be the in numbers yeah we just just could ask that very topic right now right we just finished discussing that uh is there any other question yeah hey vip and this is tom um i have two questions one is technical one is business uh the technical question is um why was fabric selected uh what was the criteria for fabric being selected versus other private blockchains like quorum or quarter or something like that and then the the business question was it's kind of an extension of what manny wrote in in the comments um and that is why a blockchain versus some other um solution um that is what problem is this actually solving or what what did you guys find um that was facilitated or made easier by using a blockchain based solution okay um let me let me answer first on the second question actually you could ask for internet but from using the centralized standardized approach continue that uh we can't hear you properly you hear me better now oh yes much better um actually uh yeah as uh as uh as you have seen as you express the the good observation on the kind of panels that that if there are any as is a person you have to stick to many panels but um we i want to tell you that we could as well implement this uh on the centralized based approach whereby the time of time and manage everything but instead we didn't choose to take that approach because we want to learn more about blockchain but um and also in in terms of setting bond sales in thailand we the there are only this many participants only for selling agent banks that are currently in the market so it's not very sophisticated um not too sophisticated to implement um uh this way so um yeah this is this is for my part so so the for so for the question that why we select the hyperledge of fabric because at the time that we start the project uh there are a few uh option platform that are on the cloud and so that uh if we if we if we select this solution so we don't have to have absorb a lot i mean like the transition is quite large number so vot don't have to uh manipulate the overall business business case because some some business case also not related to vot but we also need the data like the data that we have to create we have to create a bond uh the sorry the data that selling agent has to create a bond so actually this data vot is not it's not like it's not like you have to own this uh system by our cell but actually it's just tsd do it by themselves so but our data also um circulate into the same into the into this network so that's why we we select a hyperledge of fabric yeah just a quick uh follow up to that question was there anything else on the ibm blockchain platform that you took advantage of or in the ibm cloud that's really about at the time of can't hear you again uh sorry all right okay so at the time of our implementation of this project the commercial banks including the the awful for selling agent banks involved in this project they all uh uh developed a uh a blockchain node based on hyperledge fabric uh within the the bci blockchain community initiative um so they already have the existing knowledge and already owns Nord wish they they can utilize on our project as well so that's why um we also talk to them and we also learn from from the bci projects so that um we kind of adopt the uh the idea of using hyperledge fabric as well okay thank you so it looks like Mohan has got a question hey weapon hey thanks for this presentation you know i am more curious about see i read a lot about d5 right the decentralized finance is the possible right so in your implementation i'm trying to figure out where exactly does the decentralization aspect of bond issuance comes into play you know as i see your solution uh i see that you're trying to drive more efficiency in in the bond transactions rather than decentralization of the bond market and the bond business process i mean i just i i'm just seeking clarification can you and i think tom asked this question in a different way so can you just um um highlight what the uh the thinking was when you went into a blockchain base even a dlt could have helped you decentralize you know a dlt distributed ledger technology so where is the consensus process where is the endorsements coming in if you can just uh in your process did i make sense to pin or if you can rephrase uh let me let me try to answer it based on what i heard from these uh people i mean from a cantita than from chavisa because that will also clarify whether they presented this uh in such a way that i could understand right i mean it's not just uh just about uh you know what approach they took but whether we understand why why they did it did it this way first of all about decentralization this is a buzzword that has been around and according to scott stornetta any amount of decentralization meaning uh right now for example if you look at all those channels like the public bond channel you can see there are many many participants there and previously with centralized systems i don't think they would have had access to those that data that easily so decentralization is not just uh about putting it in a system where there are hundreds of uh nodes thousands of nodes and all the people uh participate like in a d5 situation this is more of a c5 situation but it is more decentralized than the legacy system they showed you that is my interpretation that is fair you know that's what i call driving process efficiency um you know in the past it was hopping and there were disconnected systems but now the visibility is on a shared ledger um and at least like each one of them has a copy of it so you can at least have tamper proof copies of the same transaction but think about this it also allows for a launchpad into the next step which is creating secondary markets integrating with payment systems more efficiently doing oms directly on the blockchain uh segregating the data in a separate uh data uh you know data store like an ipfs or something um you know these kind of things are only possible if you start down this path you it is not going to be a big bang approach according to me anyway but maybe if kanthitha or or chavisa have any other points to make about this i would be glad to hear yeah in our option it's uh our project is only a small project actually only small use cases that adopt blockchain and a lot of a lot of elements involved still rely on existing networks and and it's best for the ecosystem to still be relying on the existing work but however we along with blockchain we might as well uh use the existing system while making uh frances api so that we can best utilize our existing infrastructure as much as possible so it's it's kind of what's best for the for the the whole ecosystem actually in my opinion okay sounds great uh now money has a question looks like uh just to follow up on what vipin has already raised what are your next steps um you know beyond beyond what you've done so far and also are there plans to integrate with your cvdc um well this is here pretty much focusing on bond first and then um this in this case you only focus on the saving bonds that are issued in partner market we are planning to extend the usage uh linkage to the other kinds of bonds especially uh government bonds that are issued via the ecosystem and also we are looking for the possibility of linking the uh so with other systems so that we market transactions and also to link with the the um informations that are that are on the corporate bonds side that kind of plan i can get all the informations like both the government bonds and public bonds and then private bonds so that's the the way we work on the on the kind of um the implementation that we are looking to play for next year is to uh on it's basically focusing on the government bonds that are offered in in private market and then reinmarket later on so there's a lot of um here to work yes a lot a lot to be done but but with cvdc we still have yet to to um to um we're not sure whether it's a it's useful available with the uh adopted in in our network so um the payment side is still handled uh externally right with the with your um with the payments yes with the rtgs system yes um so rtgs system that means that the payments are not direct they are going to the selling agents who then pass on that uh payment individually to the individual investors um yeah kind of but the platform also facilitates payments a little bit in terms of it generates the reference uh reference number for the selling agent banks so that once the selling agent bank transfer uh money through the rtgs system then the the system the payment system will then acknowledge that this transaction is made for uh saving bonds yes but how does the money reach the investor like uh you know person who has put in let us say 300 bath or 100 bath or whatever whatever the you know so the selling agent pretty much manage the the fund transfer so they collect funds from the investor then uh they transfer money to the bank of thailand then the bank of thailand to the PDMO oh okay that is one way and now the other way when the payment is uh when the bond coupons are paid but you haven't yet integrated the coupon payment or anything into that system right no no no we have not are there plans for doing that uh we need to explore in detail first yeah sure it's a possibility yeah yeah i mean anyway so we have spent a delightful hour listening to you and asking questions this is very exciting because even though it is only disrupting a portion of the whole ecosystem it is the first step and tom if you remember the um there was a paper by martin walker which talks about the hybrid way in which blockchain will appear that it's not going to be a big bang approach and then parts of the system will get disrupted first and then uh slowly the other parts will be integrated into the blockchain yeah and i think uh in general this this is a point i often make is that blockchain at this point really is just plumbing and whether i do it with a central database or i do it with a distributed ledger almost doesn't matter what matters is the solution that i build on top of it um and really at some point we're going to stop talking about blockchain because it just becomes another solution pattern that's true but the blockchain does bring that uh that limited decentralization we were talking about before which is not available in the centralized database but on top of that you're right that when blockchain disappears in fact i have a article in Forbes about it which i say we will know that blockchain succeeded when we don't talk about blockchain anymore and that's just because it disappeared yes disappeared it is like tcpip or htp nobody talks about how it's implemented anyway thanks for showing up thanks it's very late there in thailand have a good night and we will interact more on the on the channels on email and i will try to get together some of these questions and as a meeting minutes and let's hope this is the start of something beautiful thank you okay thank you for having us here and we're glad to be sharing our knowledge and our experiences i hope it's informative and useful for you guys thank you thank you and bye bye