 So today I'm going to talk about Make-A-Fill. So this is a Python package that we worked on at CryptoEconVab. This package contains essentially a model of the Falcon economy. So it allows users to provide a few parameters that encode some behavior of storage onboarding and then based on those parameters it simulates a few metrics about the Falcon economy. And in particular it's focused on circulating supply and the components that make up circulating supply, which is how much fill is being minted, how much fill is vesting, how much fill is being locked and how much fill is being burned due to gas usage. So this will be a demo, so if things break, it's a demo, so it's fine. And I'm just using here a Jupyter notebook from Google Collabs, so we can share the link afterwards and anyone can run it and play with the code. And so here I'm just installing so this can be installed via pip, which is a common way for Python packages. And while it is being installed, I'll just say that this is the repository for our code. You have here more information about each parameter, what the parameter means, how to use it, a few examples, and then some references at the end. And we also have here some notebooks that you can run locally instead of just running here on the Google Collab. So it was installed and now we are just loading the package, which is Make a Fill and a few others that are relevant. And so the first step is defining the parameters for the model. So I'll go over each parameter and what the parameter means. So the first few parameters are related to time. So when do we want to run the simulation? So we need to provide the current date. So this is essentially used for collecting past data that we need to start the simulation. We can also provide a start date. And this means that you can run simulations in a date that is past the current date. And the reason for this is if we want to do some backtesting and do scenarios of what the economy would look like in the past, we can have a start date that is later than the current date. If we want to do just forecasting into the future, we need to have a start date that is at most two days before the current date. And this is because we need to collect some statistics to restart the simulation. And so once we define the start date and the current date, then we define how many days we want to forecast into the future. And so how many days will be model starting from the current date? And so here we are just starting at the start of December of this year and we are forecasting one here ahead. The next is what is the daily power being onboarded? And here we are referring to the raw byte power. And this is measured in PIPs and there's two options here. You can either give just a single number as we have here in the example. And if you give a single number, this means that for the entire forecasting period, so this year that you are forecasting in the future, you'll assume that every day this is the amount of new raw byte power being onboarded. But you can also give a vector of numbers. So instead of giving a single number, you can give a list. And here this means that you have more flexibility. So you can encode other assumptions like, for instance, imagine that the raw byte power is increasing a little bit every day or it increases and then decreases. So you can model different behaviors of onboarding for the future. But for this example, we'll start just with the simplest one, which is assume the same level of onboarding. And here we are assuming 10 PIPs per day. The next one is the renewal rate. So this is the percentage of power that is scheduled to terminate that instead of terminating renews. So when a sector is onboarded in Filecoin, it has a period of time that is being committed. And when that period runs out, the storage provider has the option of just letting the sector expire or they can renew the sector. And so this is essentially the percentage of raw byte power from the power that is scheduled to expire. How much is being renewed? So the more power that is being renewed, this means that the more power the network is retaining every day. As with the power here, the second parameter, this one can also be a number or a vector. And then we have the fill plus rate. So this is the percentage of power that is being onboarded. So from these 10 PIPs that is being onboarded every day, how much of that power includes fill plus deals? And again, this can either be a number or a vector. And here we are just assuming 10%. And finally, we have sector duration. So another choice that storage providers have is when they are onboarding the sector for how long are they committing the sector. And here, this is controlled by these duration parameters. So here you are assuming that all new sectors that are being onboarded, they are being onboarded with the duration of one year. And these can only be one value. So you cannot provide here a vector. So it's just all sectors have the same duration. So as you can see, these are all the parameters. So we simplified a lot what type of behavior we can encode in this model and to make it as simple to use and interpret as possible. But of course, this is a simplification, right? In the end, you have different storage providers coming in with different durations, different renewal rates. So this is a model that tries to simplify all that into a set of parameters that then you can tweak and run scenarios with. So once these parameters are fixed, then we can just run the model. So here we are using these run simple sim. We are putting in all the parameters we defined, and then the result comes in a data frame. And here I'm just selecting, so this data frame will include more statistics than what is being shown here, but I'm just selecting the ones that I talked about previously. So we have for each day, what is the circulating supply? What is the the file coin being minted, being vested, being locked, and then being burned? And then we can do some visualizations. So here you can see that we have here the main assumptions of how much power is being onboarded, the renewal rate and fill post rate. And then we have here all the components for circulating supplies. So the field that is being mined and vested, they contribute positively to circulating supply. So this means that they are increasing circulating supply and so they are positive. On the other hand, the field that is being locked or burned, they reduce circulating supply and so they contribute negatively to the final circulating supply. And so we have here, if we add the mine invested and then we subtract, locked and burned, we get this value for circulating supply for one year ahead. And this is all that is to it. You can then use different functions to just model specific parts. So if you just want to see the power being on board there, or if you just want to see the field that is being mined, we have specific functions. So then if you don't want to run the full simulation, you can dig in and use just some parts of it. And you can also play with different parameters and see what would be the impact of changing specific values there. So for instance, here we have an example where instead of having a renewal rate of 40%, we assume we have perfect renewal rate. So all the sectors are renewing every time. But all the other values are the same. And here we can see that. So it's running the simulation and then building up the plots. You see that it doesn't look very different in terms of my minefield and vested field. The big change here is how much field is being locked. So again, because sectors are renewing more, they are not exiting the network. And so that field that was locked on those sectors is not being released. And so we have more field being locked than these decreases a little bit in supply. But then we can also have an example here where instead of giving a single value for the onboarding power, for the power being onboarded, we can give a vector. And here we are assuming that we start with the initial 10 pips and then we grow a little bit every day. And we can see how circulating supply would change. And you can see that when the plot shows again, here we see now a bigger change. So now we are growing our network even faster. And this means that we are locking even more field. And this means that the circulating supply then starts decreasing. Okay, and so this is what you can do with this model. You can play with different values and then see what would be the impact on the different components of circulating supply. And once again, if you want to read more about how this model is designed, the main assumptions that you have here the links. You can also go to our repo, install it, play with it. And I want to just finish by thanking Tom and Kiran. They worked a lot on this model with me. And so they are not presenting today. Kiran is here. So if you want to give a shout out, he's here as well. But they were a great help here. So thank you.