CipherShield

Inspiration

There are too many flaws in the current credit-score calculation such as monopolies of a few organizations, less privacy and security and hence we plan on to solve this using CipherShield

What it does

A decentralised credit scoring system on the blockchain that employs zero-knowledge proofs to assess users' creditworthiness without revealing detailed financial information.

How we built it

We researched and came to know how much potential zk and AI individually had as technologies and hence we found out a way of combining those using a technique known as Federated Learning.

Challenges we ran into

Finding a way on how to integrate AI with zK and pass on the data from zK to AI without revealing the information which goes against the main motto of zK. This is where Federated Learning came into picture and solved the issue by running the computations on the users device itself and not passing on the user data to anywhere else.

What's next for CipherShield

We plan on to integrate it with multiple API providers and expand the horizon of getting the data such as real estates, etc, thereby not limiting ourselves to crypto and banking history.

Built With

• ai
• express.js
• federatedlearning
• node.js
• rainbowkit
• react
• tailwindcss
• wagmi
• zk
• zksnark