Trust Vault
On-chain proof of authenticity for creative work with privacy-preserving verification.
Problem Statement
TrustVaultis a decentralized platform designed to bring transparency, authenticity, and ownership verification to the world of digital creation. In today’s age, where content can be easily copied, modified, or falsely claimed, there is a growing need for creators to have atrustless, tamper-proof systemthat can verifywho made what and when. TrustVault answers this challenge by giving creators—artists, writers, developers, designers, and innovators—a way to permanently register their creative works on the blockchain with cryptographic proof of authorship, all while maintaining privacy and control over their content.When a creator uploads a file through the TrustVault platform, it doesn’t just generate a simple hash. Instead, it produces two distinct digital fingerprints: acontent hashand asemantic fingerprint. The content hash is a cryptographic representation of the file, ensuring that even the smallest data change can be detected. However, what truly sets TrustVault apart is its integration with theArtificial Superintelligence Alliance (ASA), which powers the creation of the semantic fingerprint. Using advanced language and vision models, this AI layer analyzes themeaning and conceptual contentof the file—whether it’s text, artwork, or code—and creates an intelligent vector-based representation. This means that even if a work is paraphrased, reformatted, or slightly edited, TrustVault can still identify its semantic similarity, helping prevent plagiarism and ensuring rightful credit to the original creator.To make these proofs verifiable, scalable, and interoperable across multiple blockchain ecosystems, TrustVault integrates theAvail modular data availability layer. Avail serves as the decentralized backbone for storing the cryptographic proofs and metadata of each registered work. Its robust infrastructure ensures that the data is available, verifiable, and resistant to manipulation or censorship, even when referenced from different blockchain networks. This modular approach allows TrustVault to be chain-agnostic—meaning a creator on Ethereum, Polygon, or Solana could all anchor their proofs in the same trusted layer. Avail’s integration also helps the system remain lightweight and cost-efficient, separating data availability from execution while maintaining integrity and auditability.On the technical side, thesmart contract architectureof TrustVault, built usingSolidity and Foundry, manages proof registration, timestamping, and verification. Each registered work is given a unique ID and mapped to both its content and semantic fingerprints, along with creator details and a timestamp. This ensures that every creative proof is easily traceable and permanently stored on-chain. Thefrontend interface, hosted onLovable, is designed for simplicity and accessibility, allowing users to upload, verify, and view ownership proofs with just a few clicks—without needing deep blockchain knowledge. It communicates with the smart contracts through APIs and leverages secure wallet connections for authentication and signature-based actions.By combiningAI-powered semantic analysisandAvail’s decentralized infrastructure, TrustVault moves beyond basic hashing to create a more meaningful, intelligent proof-of-ownership ecosystem. It transforms digital intellectual property into verifiable, immutable on-chain records while keeping users in full control of their data. The platform’s future vision includes features like privacy-preserving verification using zero-knowledge proofs, creator reputation systems, and cross-chain interoperability—making it a foundational layer for a new era oftrustworthy, AI-enhanced digital creativity.Ultimately, TrustVault’s mission is to protect creators from plagiarism and unauthorized use by giving themindisputable, timestamped proofof originality and authorship—without forcing them to expose their entire work publicly. It empowers creative professionals to say,“This is mine, and I can prove it,”while promoting transparency, fairness, and authenticity in the Web3 creator economy.
Solution
TrustVaultwas built with a focus on combiningAI-powered semantic verificationandmodular blockchain infrastructureto create a robust proof-of-authenticity system for digital creators. The project is divided into two main layers —the smart contract backendandthe AI + data availability layer, all connected through a clean, intuitivefrontendhosted on Lovable.Thesmart contract backend, written inSolidityand tested usingFoundry, is the core of TrustVault. It handles the registration, verification, and timestamping of proofs. Each file uploaded by a user generates two unique identifiers — acontent hash(SHA256) and avector hash, which represents the semantic fingerprint. These are stored immutably on-chain along with metadata such as creator address, timestamp, and a reference URI. The contract emits events for every proof registration, ensuring full transparency and enabling easy querying in the future.For theAI fingerprinting layer, we integrated theArtificial Superintelligence Alliance (ASA)APIs, which use lightweight embedding models similar toMiniLMto generate vector representations of text or image content. This AI model runs client-side (via @xenova/transformers) to ensure privacy — the raw content never leaves the user’s browser. Only the generated embeddings (semantic vectors) are used to create the vector hash stored on-chain. This was a key design decision to ensuretrustless verification with privacy. It also allows semantic similarity detection — so if someone rephrases or slightly modifies content, TrustVault can still detect its origin.TheAvail integrationacts as the project’sdata availability and verification layer. Using theAvail Nexus SDK, we connected our smart contract events to Avail’s modular DA layer. Instead of storing heavy data or metadata directly on-chain, we post proof commitments and metadata references to Avail, ensuring scalability, cross-chain accessibility, and censorship resistance. Avail’s lightweight modular architecture allows TrustVault to remain interoperable with multiple ecosystems (like Polygon or Ethereum) without rewriting core logic. Essentially, the blockchain handles the proof registration, while Avail ensures that the metadata and cryptographic proofs are verifiable and permanently retrievable.On thefrontend, built usingNext.jsandTypeScript, hosted onLovable, we created a clean dashboard where users can upload content, view ownership proofs, and verify authenticity. The frontend interacts with the smart contracts viaethers.jsand theAvail Nexus SDK, while the AI fingerprinting runs directly in-browser. We implemented ahybrid compute flow: the semantic fingerprinting runs locally, the proofs are uploaded toNFT.Storage/IPFS, and the smart contract stores the hash references immutably on-chain. This separation of concerns makes the system decentralized, privacy-preserving, and efficient.One hacky but powerful element of our build was creating asemantic hash normalization step, where minor whitespace or punctuation changes don’t produce entirely different hashes. We combined both cryptographic and AI-based hashing to strike a balance betweenstrict integrity checksandsemantic similarity detection. Additionally, we implemented a temporary batch uploader that signs transactions off-chain and queues them for on-chain submission, reducing gas usage during testing.In short, TrustVault’s architecture combines:Solidity + Foundry for smart contract logic and testingAvail Nexus SDK for decentralized data availability and cross-chain verificationArtificial Superintelligence Alliance API + MiniLM embeddings for semantic AI fingerprintingIPFS/NFT.Storage for immutable storage of content metadataNext.js (Lovable) for the web interface and user interactionsEthers.js + Wagmi + WalletConnect for blockchain communicationBy combining these technologies, we created a system that is lightweight, scalable, privacy-preserving, and truly decentralized — giving creators a powerful tool to anchor their work’s authenticity on-chain while leveraging AI to understand and protect meaning, not just bytes.
Hackathon
ETHOnline 2025
2025
Contributors
- Prince-PT
19 contributions