Fusion Credit

Fusion Credit creates a credit score called the “FusionScore” based on multiple addresses and chains while keeping the account information private. Digital signatures are used to ensure account ownership, and zero knowledge proofs are used to hide the account information and ensure the credit score calculation is done correctly.

Like real world credit scores (FICO score), the FusionScore enables financial products to model risk based on an individual user’s risk profile, and it is a basic building block of any financial system. Financial products, such as loans, insurance, leveraged investment accounts, etc., can offer personalized rates and services based on the credit score, making them competitive to other products, and more compelling and capital-efficient for the end user. For the credit score to be accurate and trustworthy, it needs to encompass all user activities, hence using multiple addresses on multiple chains. However, in the blockchain space, it is common to use multiple accounts for different purposes or activities, and the user may not want to make the link between all these accounts public. Fusion Credit leverages zero knowledge proofs to hide all the account information while ensuring the ownership of these accounts, as well as that the correct calculation process, is followed to generate FusionScore off chain. Not exposing the underlying account information is the main differentiator compared to other similar products in the space.

Another unique feature of FusionScore is that the exact process to create the score is made public (built into the zero knowledge circuit). This allows the community to contribute and come up with a better methodology for a more effective score. As more DeFi products come to market, and analytics abilities improve, a new FusionScore version can be created to keep up with the latest trend. Being able to upgrade the zero knowledge proof circuit is a feature built into the Fusion Credit system. The transparency will also encourage more products to use FusionScore, integrate with Fusion Credit System, and eventually create a healthy ecosystem. Any curious individuals or organization can also verify the score. In the future, a DAO can be created to govern this process of updating the methodology of the score.
Github: https://github.com/amadeobrands/fusion-credit

Frontend: Next.js, Web3-React, SnarkJS, circom Backend: Solidity, Hardhat, SnartJS Verifier

The high level FusionScore is rule based and can be made up with the following logical components. Machine learning can be used to create individual component scores. Longevity – Amount of time since the account was created Activity – Transactions volume, frequency, type, etc. Equity – Assets and properties owned, such as account balance, tokens, NFT’s Diversity – Interactions with different types of service Specialty – Depth of experience in a particular type of service Propriety – Having good behavior and not associated with criminals Ubiquity – Using multiple accounts and L1/L2 networks, or even off chain data

In our slide deck we have a more in-depth representation of how it works and what technology is used. After this hackathon and validation of the concept, we plan to expand the logical components that we use and harden our algorithms.