⚙️Layer2 Data Rollup

Validator

At the core of the network, Validators play a critical role. They are responsible for receiving, verifying, and batching web transactions from Routers. Validators generate Zero-Knowledge (ZK) proofs to checkpoint this session data on-chain, allowing datasets to be verified for data provenance and tracked throughout their lifecycle. Initially centralized, the validator framework is transitioning to a decentralized committee to enhance trust and security.

Router

Routers are the connective tissue linking Beans Nodes with Validators. They ensure the network of nodes remains accountable and efficiently relays bandwidth. Routers are rewarded based on the total bandwidth they validate, encouraging optimal performance and reliability.

Beans Node

Beans Nodes utilize the unused bandwidth of users’ devices to scrape public web data, which is then transformed into structured datasets for AI. Operating a node is straightforward and free, with node operators compensated for the data relayed through their devices.

ZK Processor

The ZK Processor aggregates validity proofs for all web requests and submits these to a Layer 1 blockchain. This creates a permanent record of every scraping activity, laying the foundation for complete transparency in the provenance of AI training data.

Beans Data Ledger

The Beans Data Ledger serves as the crucial link between the scraped data and the Layer 1 Settlement Layer. It is an immutable data structure that hosts full datasets and links them to their respective on-chain proofs, ensuring the integrity and provenance of data.

Edge Embedding Models

Edge Embedding Models process unstructured web data into structured formats suitable for AI models. This includes cleaning, normalizing, and structuring the data to meet AI requirements.