ML-ERC46

Screenshots

Problem Statement

Sample implementation of OpenZeppelin ERC4626 contract in Stylus using Rust which aims to enable heavy AI/ML computation for DeFi projects that want to enrich their smart vaults with machine learning capabilites that are historically performed offchain.It provides a library with augmented ML capabilites thanks to the rustlearn crate, a machine learning package for Rust as seen in the rebalance function.

Solution

This project is used using the Arbitrum Stylus SDK and written in rust. We used hardhat and js to perform tests and interaction with the deployed contract. We have also implemented rustlearn crate to perform ML computations for rebalacing purpouses.This is useful for many use cases in DeFi that often compute their algorithms off chain for rebalancing or AI/ML engagement purposes. This library demonstrate that it's now possible to perform these computations on chain