AssetForge

An AI-driven RWA lending protocol using Pyth, Filecoin & ASI to unlock DeFi liquidity.

AssetForge

Created At

ETHGlobal New Delhi

Project Description

Essential RWA documents, such as deeds and appraisals, are stored and immutably verified on Filecoin, creating a transparent and tamper-proof compliance layer. The entire system is managed by an intelligent on-chain agent, built using the ASI Alliance's uAgent framework, which autonomously assesses risk, approves loans, and manages liquidations, creating a truly decentralized and scalable financial primitive.

How it's Made

Our protocol seamlessly integrates three key technologies to create a trustless and automated lending experience for Real-World Assets. The user journey and underlying mechanics are broken down into four distinct steps:

  1. Collateral Deposit & Document Verification (Filecoin) A user initiates the process by depositing an RWA NFT (e.g., representing a piece of tokenized property) into our smart contract. Simultaneously, they upload the corresponding legal proof (like a PDF deed or appraisal) via our interface. This document is instantly pinned to IPFS and its hash is stored on Filecoin, creating an immutable and verifiable on-chain link between the digital asset and its real-world legal status.

  2. Real-Time Valuation & LTV Calculation (Pyth Network) To determine borrowing power, our smart contract directly queries the Pyth Network pull oracle for a live price feed of a relevant proxy asset (e.g., a Real Estate ETF for a property RWA). This guarantees the collateral is valued against current market conditions, not stale data. This live price is then used to calculate the precise Loan-to-Value (LTV) ratio for the loan.

  3. Autonomous Loan Approval (ASI Agent) The AI Agent, built with the ASI uAgent framework, acts as the decentralized underwriter. It performs two critical checks:

It verifies the integrity of the RWA by confirming the document hash on Filecoin.

It analyzes the LTV calculated from the Pyth price feed to ensure it's within a safe, pre-defined threshold (e.g., < 70%).

If both conditions are met, the agent autonomously approves the loan, allowing the user to borrow stablecoins from the lending pool.

  1. Continuous Risk Management (ASI + Pyth) The protocol's health is not a one-time check. The ASI Agent continuously monitors the Pyth price feed post-origination. If market fluctuations cause the collateral's value to drop and the LTV to rise above the safety threshold, the agent automatically executes pre-programmed logic—such as triggering a partial liquidation or alerting the borrower to add more collateral—to protect the protocol and its lenders from incurring bad debt.
background image mobile

Join the mailing list

Get the latest news and updates