Blockchain Compliance That Runs Before Transaction Settlement
This article introduces AMTTP, a pre-settlement transaction validation system that runs parallel to Ethereum's 12-second block window. AMTTP uses a combination of machine learning, graph analysis, and cryptographic proofs to detect fraud, screen for sanctions, and ensure compliance before transactions are finalized.
Why it matters
AMTTP's pre-settlement validation approach could help blockchain-based financial services improve compliance and reduce fraud risk before transactions are finalized.
Key Points
- 1AMTTP executes multiple validation layers before transaction settlement, including fraud detection, sanctions screening, and compliance checks
- 2It uses a hybrid ML model (XGBoost + GraphSAGE) for fraud detection and Memgraph for counterparty risk scoring
- 3ZK-SNARK proofs are used to verify KYC validity on-chain without exposing personal data
- 4The system provides a deterministic outcome (Approve / Review / Escrow / Block) and a back-office dashboard for real-time monitoring
Details
AMTTP is a pre-settlement transaction validation system that runs in parallel with Ethereum's 12-second block window. It executes several validation layers before transactions are finalized, including hybrid ML-based fraud detection, sanctions screening, counterparty risk scoring, and ZK-SNARK compliance proofs. The fraud detection model combines XGBoost for transaction-level analysis and GraphSAGE for graph-based risk assessment. AMTTP also screens transactions against OFAC, EU, UN, and UK sanctions lists, and evaluates them against FCA policies. Cryptographic ZK-SNARK proofs are used to verify KYC validity on-chain without exposing personal data. The system provides a deterministic outcome (Approve / Review / Escrow / Block) and a back-office dashboard for real-time monitoring, alerts, and explainability. AMTTP is currently a proof-of-concept running on the Sepolia testnet, with a UK patent filed and a public SDK available.
No comments yet
Be the first to comment