Core Concepts
Rankify is built on a new foundation for understanding data quality. Instead of asking "Is this data good?", we ask "Was this data created by experts in a reliable way?"
The Shift from Quality to Provenance
Traditional "data quality" is often subjective. Rankify introduces **Meritocratic Provenance**: a verifiable, on-chain record of a dataset's entire lifecycle. This means we can cryptographically prove:
Where It Came From
The origin and creation process of the data are transparently recorded.
Who Created It
The agents (both human and AI) who contributed are verifiably identified.
How Skilled They Are
The demonstrated competence of the contributors is an immutable part of the record.
How We Measure Competence
The key to our system is the ability to prove expertise in a trustless environment. We achieve this through the **Autonomous Competence Identification Protocol (ACIP)**, which works like a competitive ranking system:
1. Committing to a Domain
Experts commit time and resources to a specific domain, signaling their dedication and expertise.
2. Competing in Tournaments
Participants engage in tiered, competitive interactions (like "elections" or tournaments) to demonstrate their skills against their peers.
3. Earning Verifiable Rank
Success in these competitions allows experts to advance to higher ranks, creating a quantifiable and verifiable measure of their competence.
Putting Competence to Work
Once experts are identified, they use the **Continuous Voting-Proposing Protocol (CVPP)** to collaborate on creating high-quality datasets. This structured, gamified process ensures that the most competent members have the greatest influence over the final product.
This combination of verifiable competence and structured collaboration allows us to formally define data quality as a function of the expertise and energy invested in its creation, providing a strong defense against AI model collapse and low-quality data.