5_eq_5_bwgreen_edited.jpg

ML AND DATA SCIENCE FOR WEB 3.0

We research and develop machine learning applications, publish data, and study the economics of Web 3.0 platforms:

- decentralized applications (dApps)

- nonfungible token (NFT) marketplaces

- decentralized finance (DeFi) platforms

 

MACHINE LEARNING APPS

NOTE: These apps are for DEMO purposes only and are not updated or maintained regularly.

 

This app predicts NBA Top Shot Moment valuations by serial number and generates visualizations. Owners can value their moments, decide on the best price to sell, or determine if a moment for sale is listed at a fair price.

nba_demo.PNG

Collectors buy and sell blockchain-based cards representing football (soccer) players in an application called Sorare. Some cards have sold for thousands of dollars.

This app generates automatic valuations for digital collectibles like Sorare player cards. Owners can value their portfolios in real-time and determine reasonable prices for buying or selling different cards. It leads to more active trading by democratizing information and reduces the risk that collectors do not make the most of their investment in a digital item.

Sorare_demo.PNG

Axie Infinity is an online game where players acquire virtual assets called Axies, which they actively trade on a marketplace. Determining the value of an Axie requires a lot of experience with the game and significant time spent scanning the market for comparable assets.

This app produces automatic and real-time valuations for in-game digital assets like Axies. Players can value their portfolio at any point in time, find a reasonable selling price for their assets, or decide if an asset they wish to purchase is priced fairly. It reduce the barriers to entry for new players, immediately empowering them with data.

Axie_demo.PNG

SuperRare is a marketplace for trading blockchain-based digital art.

This app produces automatic and real-time valuations for digital art. Collectors can value their portfolio at any point in time, find a reasonable selling price for their assets, or decide if an asset they wish to purchase is priced fairly. It reduce the barriers to entry for new collectors, immediately empowering them with data.

SuperRare_demo.PNG

Participants have spent millions of dollars buying and selling virtual parcels of land in Decentraland, a virtual world built on the Ethereum blockchain.

This app generates valuations for a user-input land parcel. It helps sellers find the right price to sell their land and helps buyers determine if a land parcel is priced reasonably.

DL_demo.PNG

Participants have spent millions of dollars buying and selling virtual parcels of land in Decentraland, a virtual world built on the Ethereum blockchain.

This app generates demand curves for a user-input land parcel, showing the probability that a sale will succeed as a function of the land's sale price. It helps sellers find the best price to sell their land given their risk tolerance and helps buyers determine if a land parcel is priced reasonably.

DL_price_demo.PNG

This app visualizes the correlation between any pair of cryptocurrencies. Currently, it is one of the most flexible correlation calculators that allows for a wide variety of input tokens.

Correlations are important for liquidity-mining (assets that are highly correlated are less likely to generate impermanent loss for liquidity providers in automated money markets like Uniswap).


Similarly, understanding correlations can help reduce risk in cryptocurrency portfolios.

correl_demo.PNG
logo.jpg

DATA AND RESOURCES

​​The demos on this site use the following resources:


NFT projects analyzed:

Additional data sources:

 

CONTACT US

Please reach out to Pavel Kireyev for any comments, potential collaborations, or simply to exchange ideas. Please note that we may not be able to reply to all mail.  

pavel.kireyev@insead.edu

www.pavelkireyev.com

LinkedIn

Current Interests: Applications of data science to Web 3.0 - decentralized apps (dApps), nonfungible tokens (NFT), and decentralized finance (DeFi), as well as data portability, crowdsourcing, and platform design in Web 2.0 markets.

Background: Pavel Kireyev studies digital platforms and teaches AI Strategy (MBA) and applied machine learning (PhD) at INSEAD. He led ML projects at QuantCo, helping the company grow from pre-incorporation to 80+ employees, and worked in Tokyo, Japan and the US. Before that he was a grad student at Harvard and did applied econometrics and game theory research.