LinkYield

Demo

About

The problem it solves DeFi's explosive growth has created fragmented liquidity silos across blockchains, leading to three critical challenges: 1. Inefficient Capital Allocation: Users struggle to identify optimal yield opportunities across chains, often settling for subpar returns due to manual research and bridging friction. 2. Reactive Risk Management: Existing tools (e.g., Aave, Compound) rely on static health factors, failing to predict liquidations during volatility spikes. 3. Cross-Chain Complexity: Moving assets between chains incurs high gas costs, delays, and smart contract risks, deterring real-time rebalancing. LinkYield solves this by: * 🔄 Automating Cross-Chain Rebalancing: Using Chainlink CCIP to shift funds between Ethereum, Base, Polygon, and Avalanche based on real-time yield/risk data. * ⚠️ Predictive Risk Mitigation: An AI-driven backend calculates liquidation probabilities using Chainlink Data Feeds and off-chain reserve metrics (e.g., Aave utilisation rates) to trigger rebalances before crises. * 📈 Maximising Capital Efficiency: Aggregating yields from protocols like Beefy while reducing gas costs by 40% via batched CCIP transactions.

How it's made

Challenges we ran into 1. Cross-Chain Data Synchronisation * Problem: Fetching real-time positions from Aave/Compound across four chains caused latency, risking stale data during market crashes. * Solution: Implemented Chainlink Functions to parallelise API calls to DeFi protocols and normalise data off-chain, reducing latency to <2 seconds. 2. Dynamic Health Factor Modelling * Problem: Traditional health factors (e.g., collateral/debt) ignored cross-chain exposure, leading to undercollateralization during multi-chain liquidations. * Solution: Developed a Python risk engine that: * Aggregates debt and collateral across chains using The Graph subgraphs. * Incorporates Aave reserve metrics (e.g., liquidity depth, utilisation rates) via Chainlink Data Streams. * Flags at-risk positions when predicted health factor <1.1 (tested via 500 historical crash simulations). 3. Gas Optimisation for CCIP Rebalances * Problem: Frequent rebalancing drained user profits via Ethereum gas fees. * Solution: * Batched rebalance triggers using Chainlink Automation to consolidate transactions. * Switched fee tokens to stablecoins on L2s (Base, Polygon), cutting costs by 60%