Architecture
High-Level Design
Data Creators use vllo to build synthetic datasets from private or raw on-chain data.
Validation Nodes verify quality and consistency using on-chain reputation and cryptographic proofs.
Index Layer stores dataset references and metadata on Solana for cheap, verifiable access.
Consumers (apps, AI agents, DeFi dashboards) query datasets via a simple API.
Data Creation & Validation
Users generate synthetic data locally or through vllo tooling.
Datasets are hashed, signed, and registered on-chain.
Validators run quality checks and assign trust scores.
High-quality datasets earn visibility and incentives.
Security & Privacy
No raw data is stored on-chain — only synthetic output and metadata.
Optional ZK proofs can verify data lineage without revealing sources.
Performance
Solana’s low latency and parallel runtime allow near real-time dataset updates.
Off-chain storage (e.g., Arweave/IPFS) is used for heavy payloads; Solana stores hashes and metadata.
Last updated