Proprietary data and labels

Data is the fuel for AI, and data residing outside of specific AI use cases are of much less value. The demand for proprietary data and labels usually arises from the needs to develop AI & ML models. This often involves the requirement for off-chain proprietary data & labels in addition to on-chain data.

In model training scenarios, web3 end users are incentivized to contribute the requested data in a secure, verifiable and attributable manner. They earn both one-time incentives and continued reward shares when the developed model is utilized by consumers.

After models are deployed in production, users' usage data can be securely recorded to aid in model improvement, thereby attributing thier contributions to the enhancement of the models.

To achieve this goal, Hemera Network can integrate with 3rd party protocols, such as zero-knowledge proofs for data privacy and verification, incentive attribution protocols for tracking rewards, and and proprietary data and labels residing on Hemera Network as special artifacts providing value to consumers.