Skip to main content
TL;DR
Codatta turns human/agent contributions into verifiable, ownable data assets with traceable lineage, policy-gated & observable access, provable ownership, and deterministic royalty payouts.

​
Architecture

​
The 4-step flow (end-to-end)

  1. Contribution: A contributor (human or agent) submits work that is sealed as a Contribution Fingerprint and bound to Identity and Reputation.
  2. Assetification: Atomic contributions (samples, labels, validations) with verified contribuiton fingerprints, are assembled and curated into versioned datasets (Lineage & Asset Assembly), and are minted into on-chain Tokenized Ownership with transferable fractions.
  3. Usage: AI builders access data through the Access Gateway backed by secure Storage / Compute / Serving. Every read produces usage & billing events (metering).
  4. Royalty: The Royalty Engine combines revenue, time-pinned ownership fractions, and metering events to produce deterministic payouts to ownership holders and the treasury.

​
Key components at a glance

  • Contribution Fingerprint (CF): atomic, signed record of who did what, when, with what evidence.
  • Identity & Reputation: privacy-preserving identity plus staking-as-confidence and accuracy history to score trust.
  • Lineage & Asset Assembly: composes CFs into traceable datasets with versioning and provenance queries.
  • Storage / Compute / Serving: source-of-truth storage, secure compute (TEE/federated), and performant delivery.
  • Access Control & Metering: policy-gated access; emits auditable usage & billing events.
  • Tokenized Ownership: on-chain, transferable ownership fractions over units/datasets with proof exports.
  • Royalty Engine: converts revenue into transparent, replayable splits for contributors/backers (ownership holders).

​
How to acquire ownership (short)

  • Knowledge Contributors (providers / validators): contribute samples / labels / validations. When your contribution is accepted and published as a CF, a default share is minted to you (policy-defined per task).
  • Knowledge Backers (stakers / underwriters / buyers):
    • Stake-as-confidence / Underwrite to acquire or boost fractions; audits can increase or slash positions within governed caps.
    • Buy fractions on a secondary market (if liquidity is enabled). This does not change payout math—only who receives the owner share.
  • Transfer / Escrow: move fractions directly or place them in escrow until conditions are met (payment, KYC, milestones, disputes). Snapshots use the owner-of-record at time t for payouts.
See /core-concepts/tokenized-ownership-proofs for details.

​
Where to start

  • Building the stack: start with Contribution Fingerprint → Identity + Reputation → Lineage → Storage/Serving → Access & Metering → Ownership Proofs → Royalty Engine.
  • Exploring products: see Products → Lineage Explorer and related dashboards for UI/API usage and SDKs.
Main goal: use blockchain to turn data into an ownable asset class for AI—making rights provable, ownership transferable, and revenue shareable via royalties.