AgentVault AI is an AI-powered Web3 intelligence platform that helps users analyze wallets, crypto projects, and websites. It converts raw blockchain and project data into structured risk reports, the
AgentVault AI is an AI x Web3 intelligence platform that analyzes wallets, crypto projects, and websites, then generates structured risk and behavior reports. Users can enter an address or project, generate an AI-powered intelligence summary, and store the final report on 0G Storage testnet with verifiable proof links.
The goal is to make Web3 due diligence easier by turning raw on-chain activity into readable insights such as transaction behavior, activity patterns, risk level, key observations, and recommendations.
During the hackathon, we built the full MVP end to end:
Developed a React/Vite frontend with report generation, dashboard, history, and storage proof views.
Built a Django REST backend with endpoints for analysis, report storage, history, and individual report retrieval.
Integrated AI report generation using an OpenAI-compatible provider with fallback behavior for demo reliability.
Added wallet analysis using Etherscan V2 indexed transaction data, including normal transactions, internal transactions, ERC-20 transfers, counterparties, token activity, and recent wallet behavior.
Integrated 0G Storage testnet using the Python SDK to upload generated reports and return root hash, transaction hash, and explorer links.
Added local report history so stored reports can be viewed after generation.
Deployed the frontend and backend for testing/demo use.
AgentVault AI is currently self-funded and was built as a hackathon MVP. We have not raised external funding yet. Our current focus is validating the product concept, improving the wallet intelligence engine, strengthening decentralized storage retrieval, and preparing the project for future partnerships or grant opportunities.