Licen

AI companies are desperate for high-quality data to train their models, but independent creators and researchers are hesitant to share their premium datasets because they lose control of the data and rarely see any financial return. LICEN solves this by providing an end-to-end decentralized protocol that lets data owners securely publish encrypted datasets, define unbreakable usage policies, and earn automated royalties when AI researchers use their data for model fine-tuning.

LICEN is built natively to showcase the full power of the 0G ecosystem. When a creator uploads a dataset, it is encrypted client-side and anchored to 0G Storage via a permanent Merkle-root identity. The creator then defines pricing and access rules in the DataPolicy smart contract on 0G Chain. When a researcher wants to train a model, they pay the required escrow upfront. LICEN’s orchestrator then securely dispatches the job to 0G Compute, where the dataset is unsealed exclusively inside a hardware-isolated Trusted Execution Environment (TEE).

Because the raw dataset never leaves the secure enclave, the buyer gets their trained AI model, the original data remains protected from theft or scraping, and the smart contract instantly settles royalties directly to the data creator's wallet. LICEN replaces slow, legally murky PDF licenses with enforceable cryptography and automated micro-payments.

During the hackathon, we built a complete, end-to-end decentralized application from scratch, fully integrating the 0G ecosystem. Our completed progress includes:

• Smart Contracts (0G Chain): Wrote, tested, and deployed the DataPolicy smart contract to the 0G Galileo Testnet. It successfully handles policy enforcement, escrow locking, and automated royalty settlements.

• Encrypted Storage (0G Storage): Implemented client-side AES encryption and ECIES key sealing so that datasets are securely uploaded directly to 0G Storage using Merkle-root identities.

• 0G Compute Orchestrator: Built a custom background Node.js worker that monitors on-chain events, unseals decryption keys, and automatically dispatches secure fine-tuning jobs to the 0G Compute network. We also built a simulated fallback mode to ensure the live demo pipeline (from payment to royalty settlement) functions smoothly even if the physical 0G Compute nodes experience downtime.

• Real-time Indexing: Set up an Envio HyperIndex to track all contract events, allowing the marketplace UI to load orders of magnitude faster than raw RPC polling.

• Full-Stack UI: Designed and shipped a Next.js frontend featuring a Publisher Dashboard (for tracking royalties/active jobs) and a Researcher Marketplace (for purchasing secure training epochs).

pre-raising