# CertifiedData

CertifiedData is a platform for generating Certified Synthetic Data and issuing machine-verifiable certification artifacts for datasets and other AI artifacts.

## Machine-readable resources

/api/definitions          - Canonical term definitions (JSON, 24h revalidation)
/api/definitions/{term}   - Per-term definition lookup (JSON)
/glossary                 - Human-readable definitions (DefinedTermSet JSON-LD)

## Proof surfaces

/registry                 - Public artifact index (CollectionPage JSON-LD)
/registry/{slug}          - Individual artifact proof record
/certificate/{id}         - Certificate proof object (CreativeWork JSON-LD)
/verify                   - Verification process and tool (HowTo + FAQPage JSON-LD)
/transparency             - Immutable public audit trail

## Verification

Public key: /.well-known/signing-keys.json
Hash algorithm: SHA-256
Signature algorithm: Ed25519
Certificate format: JSON (cert.v2 - see /api/definitions/ai-artifact-certificate)

## Product

- /                     - Platform overview and certification-first positioning
- /pricing              - Free, Build, Trust, and Govern plans
- /generate             - Generate a synthetic dataset, no raw data required
- /dataset-marketplace  - Pre-built certified synthetic datasets across industries
- /how-it-works         - Full Generate -> Certify -> Verify -> Govern workflow

## Registry

- /registry                    - Browse all certified AI artifacts with SHA-256 fingerprints and verification links
- /registry/synthetic-datasets - All certified synthetic datasets
- /registry/finance            - Certified synthetic financial datasets
- /registry/healthcare         - Certified synthetic healthcare datasets
- /registry/fraud-detection    - Certified synthetic fraud and AML datasets

## Trust and governance

- /trust-center                   - Security posture, cryptographic primitives, key management, and responsible disclosure
- /certificate-policy             - What CertifiedData certifies, issuance criteria, and revocation policy
- /trust                          - How CertifiedData establishes cryptographic trust for synthetic datasets
- /ai-governance-framework        - AI governance framework, certificate lifecycle, and EU AI Act alignment
- /security                       - Ed25519 key management, auth, CSRF, rate limiting, and audit logging
- /status                         - Operational status of certificate issuance, verification, and registry
- /ai-training-data-certification - Certification for AI training datasets
- /aibom                          - AI Bill of Materials and artifact provenance for AI systems

## API and verification resources

- /docs/api               - Human-readable API reference
- /openapi.json           - OpenAPI 3.1 specification
- /docs/verification-spec - Independent verification guide
- /docs/certificate-spec  - Certificate schema reference
- /docs/signing-keys      - Signing key lifecycle and distribution

## Agent Commerce

Infrastructure for AI agents to transact, settle, and produce verifiable financial records.

- /agent-commerce                                  - Agent Commerce module overview
- /agent-commerce/verifiable-agent-transactions    - Policy-governed, receipt-issued payments architecture
- /agent-commerce/what-is-an-ai-payment-agent      - Definition: autonomous payment agent
- /agent-commerce/authorization                    - How spend requests are evaluated before execution
- /agent-commerce/receipt                          - Ed25519-signed receipt structure
- /agent-commerce/payment-verification             - Independent receipt verification
- /agent-commerce/policy-engine                    - Runtime policy enforcement engine
- /agent-commerce/spend-governance                 - Spend governance framework
- /agent-commerce/pricing                          - Agent Commerce pricing
- /agent-commerce/features/spending-limits         - Per-transaction and daily spend limits
- /agent-commerce/features/policy-gated-transactions - Policy evaluation pipeline
- /agent-commerce/rails/stripe                     - Stripe rail (Phase 1 — active)
- /agent-commerce/rails/usdc-base                  - USDC on Base rail (Phase 2)
- /agent-commerce/rails/eth                        - ETH rail (Phase 2)
- /agent-commerce/use-cases/api-payments           - API access payment use case
- /agent-commerce/use-cases/procurement            - Autonomous procurement use case
- /agent-commerce/use-cases/dataset-purchases      - Dataset purchase use case
- /agent-commerce/use-cases/saas-renewals          - SaaS renewal use case
- /agent-commerce/docs/glossary                    - Agent Commerce term definitions
- /agent-commerce/docs/event-taxonomy              - Canonical event types
- /agent-commerce/docs/schema-registry             - Receipt and payload schemas
- /agent-commerce/docs/status-enums                - Authorization and transaction status codes
- /agent-commerce/x402-comparison                  - x402 protocol vs verifiable agent transactions
- /agent-commerce/stripe-vs-stablecoin             - Rail comparison: Stripe vs stablecoin

Agent Commerce verification:
Public key: /.well-known/certifieddata-public-key.pem
Verify receipt: GET /api/payments/verify/:receiptId

## Open source

CertifiedData publishes three public repositories:

certifieddata-public (verification toolkit):
  https://github.com/certifieddata/certifieddata-public
  - Certificate verification SDK
  - Schema tools and type definitions
  - Examples and fixtures

certifieddata-agent-commerce-public (Agent Commerce contracts + SDKs):
  https://github.com/certifieddata/certifieddata-agent-commerce-public
  - OpenAPI contract
  - AsyncAPI event schema
  - Python and TypeScript SDKs
  - Mock server for local development
  - Test vectors for Ed25519 receipt verification

pii-scan (data safety utility):
  https://github.com/certifieddata/pii-scan
  - Local-first PII detection for datasets
  - Pre-certification scan integration

Open source landing page: /open-source