Strangers United

Screenshots

Problem Statement

Our dApp that allows people to see the chances of having a DeFi loan liquidated on any given day as well as see the relationships in price movements across tokens. We have include an innovative technique for fitting historical data into tiny probability distributions called 'chance-data' that may be used by smart contracts, trading bots, AI models and even spreadsheets. Chance-data has a special property that normally probabilities do not have when used in decentralized environments- they can roll-up risks and opportunities. Decentralized simulations historically are not additive. These simulations are that means they may be audited and used to rollup risks. This is a way for protocols to show systemic risk too and get ahead of any regulators. Special note we are 5 strangers the youngest 15 years old. We are located in 4 different countries and 5 different time zones. The entire project was done without any verbal communication, no set meeting times, and no burn down task lists. Good fun.

Solution

We used: React, Alchemy, Chainlink, Conjure, Covalent and IPFS. The key innovation was the fitting of historical price changes into tiny sharable json files called chance-data. We used a relatively new statistical innovation called metalogs to fit the data and compress it into a tiny files. A data pipeline creates chance-data and posted it to IPFS. There any dApp may access it and use the hydrate.js to create simulation trials that are always the same and auditable.