Flare Workshop Bring Any Web2 Data On Chain I Victor Munoz

[Music] Um hello um my name is Victor Moz. I'm a deil at Flare and today I have the pleasure to introduce a demonstration on how you can bring web two data on chain uh with Flare. uh we get a lot of questions about what is flare and what makes flare unique and what are the kinds of applications that we can build on top of flare. So the purpose of this workshop is one give you like a quick rundown about flare uh what is it all about and what you can do uh with a specific examples of applications and then we're going to have a sort of live demo. We are going to run a demonstration of a flow that will show you how you can grab data from a web two API uh verify that on flare and then perform a specific set of actions on chain based on that and this sort of flow is something that you can apply to many different types of applications which is of course kind of up to you what you can do.

So uh in that sense what you can expect for today is understanding what makes flare unique uh learning the basics of the flare data connector which is basically an enshrined oracle that allows us to verify uh web two data on flare uh and also where to find the best information to start uh building. And uh by the way, we have uh quite a few of these that will give you all that information right off the bat if you want to check it out. Uh so okay, the first question that I get from people in these kind of events is what is Flur all about? And what we answer is Flur is a L1 blockchain that is EVM compatible. And that doesn't really tell you much until I share that we have uh data protocols that are native to the chain that are enshrined right uh so if you think about for instance building on ethereum if you need any source of external data you will need to call oracles and that comes with a different set of assumptions regarding the uh security but on flare uh oracles are part of the network and that means that they inherit the same uh this level of decentralization and security.

But that's not all because utilizing these protocols, there are ways in which you can uh bring data and build solutions on top of that. So what you can see on the slide uh is uh an example of three different systems or protocols uh that you can utilize uh today and build on flare and qualify for the bounties that we have. So the first one and pretty simple to kind of explain if is the FTSO or flirt time series oracles that you see on the center it is basically an enshrined oracle that provides uh decentralized data fits uh to the uh flare network. People use this for price feeds, right? So, uh nothing too, let's say, impressive about that.

But wait a minute, because then we have another uh protocol called the FTC or Flare data connector that allows you to uh bring data from web two and other blockchains to Flare, verify that and then, you know, trigger some actions on chain. And this is precisely what we are going to explore today with a web two API example. Now the most interesting thing is that when you combine these technologies there is a a wide variety of systems that you can build on top of that and things are getting more and more exciting I would say on flare because of this with the example of F assets. So in simple terms F assets is a trustless overcolaterized asset representation for uh non smart contract platforms. So think about for instance uh Bitcoin that doesn't have uh programmability, right?

Uh what you can do utilizing F assets is essentially a user could lock their Bitcoin and utilize a minting on Flur that will connect these uh uh protocols via the FTC and FTSO so that you can have access to EVM D5. So now all of these assets that are from other chains that are sitting idle they can have access to EVM D5 and we are working uh towards implementing this with XRP and this has brought the word or the movement of XRP in which there is uh a bunch of people and a bunch of developers building solutions that are going to work for XRP and they will have like access to D5 EVM and there is a lot of capital that is going to be moved because of that. Uh so you can see how you can start from very basic uh features from these oracles and then build you know more sophisticated systems and I hope that today I can you know give you a glimpse about this and how you can actually uh implement your own solutions. Uh so today we're going to talk about the flare data connector. Um if you're curious you can uh dive in right away in our documentation.

I'm going to show you around as well. Uh so no worries about that. Uh what I want you to understand about the flare data connector is that you can uh uh you can perform atest stations of different types and that means that you can graph data from uh web two APIs and verify it on chain but you can also combine this with uh verification of actions that happens on other chains. Right now if you think about uh I'm I'm performing actions on chain A and chain B. How can I connect this and bring some logic into uh another ecosystem?

It is really a complicated uh problem to solve and we are providing a solution via these uh enshrined oracles. Now the number one question that I get when I'm explaining about the flirt data connector is okay uh sounds good in in in paper but in practic in practice what can I build with this and there is a wide variety of applications that uh you can think of in which for instance for today we're going to focus on sort of web two applications uh but there are also like web three sort of connecting with web two applications as well. But for the web two side, we can think about insurance. So imagine that you're a farmer, you have a crop and you have an insurance uh bad things happen, the weather is uh not good and there is a flood, then you would have to go with your insurance uh get a person to verify that and go back and forth before settling any sort of insurance. When uh utilizing this kind of tool, what you can do is simply fetch data from uh weather uh or maybe satellite images or combine both sources, right?

And you know, essentially settle the payment based on the fact that the data is verifying uh that essentially there was a flood on a specific region, right? And this can apply for all sorts of different uh insurance policy and you can apply this to other kinds of market. For instance, uh carbon credit. You can emit carbon credit tokens, for instance, when you verify that uh company A is producing a certain amount of clean energy. And things get really uh much more exciting, at least for me, when it comes to social media and uh systems that you know will allow you to have reputation on chain, but also they bring like web two data, right?

because essentially you can combine those sources of data to have a system that will verify uh information and then trigger some logic. Let's talk about for instance in crypto they talk a lot about Kos how effective are they? Well, you know there are a couple of things that you need to look at. You need to look at for instance what they are posting on Twitter and the quality there then what they are doing on chain and verifying those kind of things can be really really kind of complicated and time consuming. So with the flare data connector in principle you can build solutions that allow you to verify all of that and then trigger some logic which in the case of I don't know the k would be maybe a payment right so if he do like if he performs a certain amount of posts and has certain uh it is about certain threshold of reputation then he would get paid x amount that kind of thing.

Uh so yeah today we are going to explore a concrete example uh in which what I really want you to take if you're thinking about bringing web two data on chain is the general flow of the flur data connector and I can tell you right away that is certainly not possible that I can explain that to you fully in details in 10 or 15 or or 20 minutes. Uh so what I'm going to do today is I'm going to explain you the general flow such that if you are serious about this and want to get started you know where to go and what do you need to change in order to make it work because we have a starter kit uh on hard hat and also on fondrid that you can utilize to get started with this uh with a complete example and the full flow. So what we are going to be doing today is we are going to grab data from a web two API from a Star Wars place. uh we are going to perform an API call right and then perform uh some processing from uh this data utilizing a jq filter and then verifying that through the FTC server uh and you will need to prepare an attestation request and validate that and then this will be broadcasted to the network so that the validators of the network will have to vote uh and perform the fetching of the data and apply the same rules that you share on the script uh about how to go about the this data and then if it's valid if it's if there is consensus there will be a proof that will be generated right that you can then retrieve from the data application layer and then you will have your data that is verified and you will have a proof. So now here's when the onchain logic comes and you can in this case we're going to do something very simple.

We have a list of characters and we are just going to create a list that performs a calculation based on the properties of this character and adds this to uh the list and then storage everything on chain. Right? So you will go from an initial web two uh set of data that is raw to a final list that has spe special uh components that has been verified and that has been enhanced on chain. Uh and that's sort of what we are going to be exploring today. Uh so now we are going to go into the details of this demo.

It's called web tojson. Um the uh QR code that you can see uh on the center of the screen will take you to our developer guide uh which is something that I really really recommend you to get started with. Uh and on the right you also have access to the Flare Foundation GitHub. Uh there you will find all of our repositories including the ones that are going to be relevant for you which is like the hard hat started kit which we are going to be using today and the Foundry started kit. you know you can utilize whatever you want right uh and also on the left I prepared a sort of FTC 101 because when I was studying how to explain this there are a lot of details that you know it's not easy to cover and we are actually going to use the help of AI to cover some of the detail to help us explain pieces of the code uh we're going to use cursor with a project setup that is going to be the hard started kit but you probably need additional information if you want to change that to your specific application.

Let's say you want to bring data from Twitter, then of course you will need to change things on this uh project, right? So how you go about it? Well, I'm giving you some uh pointers in this documentation and I'm going to show you around. So uh okay, let's just first uh get a few things straight. So if you go to the uh Flare developer hub, you will find here how to get started building on Flare.

Like I said, it's an EVM chain. So, pretty straightforward in that sense. You have all the information about the network, the RPC, uh how to get started and deploy your first smart contract. Fairly simple if you have built on Ethereum, which I'm pretty sure all of you have. Uh and then it it becomes uh more and more interesting as you explore our different products.

So like I said today I'm going to explain uh an example of the FDC that we have here but there are also other protocols that you can utilize and you can build uh for this hackathon right so you have uh F assets as well and the FTC so we are going to go here to the hard hat guys and what we are going to be exploring today is this web two JSON uh example now uh before I go into this uh uh I want you to perform the following action so that you can follow exactly what I'm doing here and we're all on the same page. So if you go to the GitHub of Flur, you can click here on the Flare uh hard hat starter, right? And uh it has some very simple instructions for you to um just test it, right? And it's fairly simple. You uh you know you you just follow these instructions.

Yeah. Right. you you do the git clone yarn uh set up your m file the usual stuff right so if you do that then uh I am going to open up how things would look like and what we can expect from the web to JSON example so what I have here on the screen is essentially uh the web 2JSON example I open this up utilizing cursor why because uh what I want to do is not only run this example but I I want I want the uh AI agent to explain to me the different pieces of the flow so that when you have to build you know exactly what to change and which you know pieces of the code are relevant for you. Uh I have about uh five more minutes so let me get started by running the example. So you do yarn had run scripts FDC example web tojson.

ts s network custom 2 and you hit enter. Hopefully uh this is going to work. I mean we have tested throughfully and it has worked before. So no issues and then what you will see is a bunch of logs uh that might get you confused the first time, right? Uh so what is happening here is essentially we have the full flow implemented on one script uh which is on this uh folder scripts web2json.

ts. TS right that what it does is uh it will prepare an attistation request that will be sent to the FTC verifier server. Uh you will perform an AVI encoded request and then this will uh be broadcasted to the network so that the validators will vote to perform exactly the same action as you did. So you go to the uh web to uh sorry to the API of the Star Wars. You grab some list of characters and you will tell him what information you want to process and send that right and then the validators will replicate that and will vote reach a consensus and based on that what they will have there is essentially a proof that what you are sending from the web to site is valid right and they in that way the data is going to be uh verified.

Uh so just a couple more of of words about this. If you want to check out uh the API, you can do it here as at this website as swappy.info. And another thing that you might want to do is play with the jq filter so that you know, you know, what are the rules for processing this data if you have never done this before. I I have tested out like very at a superficial level to play if you can do this this with AI and you you can pretty much do this with with agents right and and cursor and and and whatnot.

Now uh what I want to stress out now is uh the in the final minutes that I have uh I want to do uh a sort of reverse engineering uh thing in which uh we take the logs that we see on the screen and we ask cursor to explain to us uh which thing happened on each step. So uh the whole process I I I run this and I I leave it there because it takes about uh two and a half minutes more or less to complete. That means uh the whole process from gathering the data and getting this verified and then the contract implemented so that you get an enhanced list with information from this character will take about that time. And the final log that you can see here is the information about one of these characters uh that has been verified onchain. And we have added a new component here which is called the BMI body to mass uh ratio or index whatever which is uh an onchain calculation that we are doing right.

Uh so now let's uh and this is the end of the process uh after we have prepared the atestation request and we have gone through uh the multiple steps uh from the flow of the FTC example. Uh I'm not going to go into the details of all of these. I'm just going to point out what is important for you and what how you can make sense of each one of these logs. And this is precisely what I did with cursor. uh I ask it okay I run this this is my log I need you to explain to me step by step so that I can then modify this and utilize it with my own example and this is the uh output here from uh this conversation it took a little bit it wasn't that complicated though and then it explains uh to me the first step is to uh request the preparation and perform the AVI encoding and is uh sharing here what is like the terminal output that I'm seeing here uh how Is this related to the lines of code from the main script at the web to JSON here?

And as you can see on this piece of code, uh we have all of the information about uh the API URL, the rules for pre processing this data. Now, what I would do is you guys have to first think about the solution you want to implement before you start gathering data and start filtering that and start creating your onchain uh solution, right? It's better to take some time to think about this because otherwise you're gonna you know kind of waste time processing data that you might not need right uh so what's happening here under the hood is uh we are defining uh the request parameters uh like I mentioned uh we have the API URL the uh postprocess jq filter you can play with this on on your own and then more importantly the AI signature that defines the expected data structure for your smart contract. Remember you are gathering data from a web two source and at the end of the day what you want is that once it has been verified uh the contract needs to know what information it's going to receive right um okay and then okay you can go over the details of uh the uh you know how you perform the request how you perform the jq filter uh and what is the response from the verifier right so the verifier will send you a log a log saying okay everything is okay. What is the status?

Uh where is your ADI encode request? And then uh you will move forward to submitting this request to the FDC hub contract. So we have a contract that takes care of managing all of the things under the hood and you don't have to modify anything here. Uh but I'm just you know letting you know what is happening so that you understand how you can implementate how you can implement this uh for uh your own solution. So, and just uh to wrap it up, I'm just going to go high level over each one of the steps.

Probably you will have questions, but I will be there. My colleague Nick will be uh on the boot as well, and we can talk about this in detail for uh the specific solution that you might want to implement. So, step two, uh step two, you submit your request to the FTC hub. Uh the steps uh you will uh do the uh the request submission. Uh then you will have a round calculation and then uh a b a a voting round assignment and then you will go over the process in which the validators are going to fetch this data vote on this and and send a proof, right?

And that proof you can retrieve uh from the DA layer. Uh and as you can see the terminal output here uh is going to share information about the uh request that has been prepared. where is the bolting round uh and in the information uh about the proof right which is what you can see uh over here. Uh okay so now you have the proof and then what is next what is the next step in in this uh is essentially you will deploy your smart contract at this point you have all of your web two data the processing rules and the verification all of that is ready then what is the onchain logic uh that you want to implement you have to think about that and at this in this flow we already have a contract that uh is going to uh be deployed and uh is very like kind of simple what it does because what it's going to be doing is just grabbing some of these uh parameters and performing a calculation of the VMI, right? Um and finally like I mentioned you process the proof and decode the data.

The terminal output that you can see here with the proof X and the response type uh is the same that you will find here on the logs. I know that the logs are kind of hard to uh see from here. So that's why I wanted to you know separate this uh with this more uh I would say easy to understand output and what is happened behind the scenes is uh you will have uh the uh response extraction and you will have the data structure definition that again uh this is something that you have to define before running all of this. uh and final step uh you decode the proof data uh and you get uh this uh terminal output here that has all of the information with the hash and the information from uh your request and the AVI encoded data right uh so there is at this step there is an AVI uh decoding uh and finally you have the smart contract uh interaction and uh by the end of the day like I mentioned what you are doing here is just gathering all of this information and performing this calculation here. Right?

So you perform this calculation and then you are enhancing the initial list uh from the information of your character uh with this new information. Uh I think that with this I I have to wrap it up but I'll be ready for questions uh outside or at the boot or whenever you're ready. [Music]