KYT
Sanctions
Know Your Transaction. In 100 milliseconds.
Real-time transaction screening: sanctions, PEP, adverse media, and wallet-risk attribution — returned in under 100ms with a versioned decision token.
Every transfer, Every signal scored
Sub-100ms risk scoring, p99 in production
Every inbound and outbound transfer is scored in under 100ms — fusing sender identity, beneficiary reputation, amount, corridor, and behavioural signals per FinCEN 31 CFR 1010.410(f) record-keeping requirements.
- 100ms median latency; p99 under 250ms measured in production
- Sender and beneficiary screened against Dow Jones, World-Check, OFAC, and UN lists
- Peer-group normalised amount and velocity checks against FFIEC BSA/AML thresholds
- Corridor-specific policy overlays for 180 jurisdictions
On-chain and off-chain risk, one score
WIDTH fuses Chainalysis, Elliptic, and TRM Labs wallet scores with core-banking KYC and fiat behaviour — one risk view, one audit trail satisfying FATF Rec 16 virtual-asset guidance.
- Wallet address classified as mixer, sanctioned, darknet, or scam
- 30+ blockchain networks supported (EVM and non-EVM)
- Off-chain customer identity linked to on-chain address activity
- Unified risk score across fiat and crypto rails, single decision record
12 typologies, tuned per jurisdiction
WIDTH detects structuring, layering, mule-network patterns, and APP scams against peer-group baselines — not static thresholds — aligned to FFIEC BSA/AML Manual typology guidance and Wolfsberg AML Principles §6.
- 12 built-in typologies, each tunable per jurisdiction and customer segment
- Network-graph propagation surfaces mule rings across 2nd and 3rd-degree counterparties
- Beneficiary reputation scoring reduces APP and scam exposure at payment initiation
- Models retrain from analyst dispositions — false-positive rate drops over time
Policy changes ship in hours, not sprints
Your compliance team authors rules in a visual builder; the engine executes them at <100ms. Every version is immutable and cited in the audit trail — meeting EU AI Act Art. 13 transparency requirements for high-risk systems.
- Visual rule builder — no engineering ticket required to change thresholds
- Per-jurisdiction overlays: OFAC, UN, EU consolidated lists; MAS PSN02 thresholds
- Shadow mode tests new rules on live traffic before production enforcement
- Full version history with one-click rollback; every change attributed to the author
