LinkedIn Watch on Twitch
Identity without disclosure

Attest Pipeline

Deterministic ingest pipeline for hostile KYC/AML payloads: canonicalization, SHA-based attestation, and cross-institution identity matching without creating a PII honeypot.

attestationidentitycanonicalizationkyccryptography
← All projects Watch Live (Kick) LinkedIn

Problem

  • Traditional KYC pipelines tend to centralize sensitive user data, which creates a high-value PII target and operational risk across every institution that touches it.
  • Institutions still need to detect duplicate or abusive account creation across regional banking partners without sharing raw personal data.

Constraints

  • Input payloads are hostile-by-default JSON and cannot be trusted structurally or semantically on arrival.
  • Equivalent payloads must yield equivalent attest artifacts across institutions.
  • The system must avoid using raw PII as the core matching primitive.

Approach

  • Sim wallets and simulated KYC/AML JSON payloads are generated, normalized, and canonicalized before hashing so that formatting variance does not change the resulting identity artifact.
  • Regional banking partners receive the canonicalized payload form and produce signed SHA attestations rather than exchanging raw user identity records.
  • The result is a stable, cryptographically matchable identity proof that supports cross-bank comparison while minimizing disclosure.

Why It Matters

  • This closes an anti-gaming gap: the same person attempting to open accounts across multiple institutions can be matched through attest artifacts without each bank becoming a long-term PII warehouse.
  • The design shifts the system from storing everything forever toward proving enough, disclosing less, and verifying deterministically.

Result

  • KYC decisioning can operate on deterministic attest outputs instead of raw identity copies.
  • Cross-bank identity matching becomes possible without turning the platform into a centralized honeypot of sensitive personal data.