Blits.ai

Blits.ai — A continuously-built cross-institution money-mule graph plus the investigator dispositions on top of it. Nodes are accounts, customers, devices, IP addresses, postal addresses and phone numbers; edges are payment flows, updated continuously from onboarding and payment data. Enriched with external signals the vendor ingests: scam reports from other banks, confirmation-of-payee mismatches, industry and central-bank mule lists, and law-enforcement requests. The valuable residue is the labelled outcome layer: per-account mule-likeness scores with the behavioural reasons attached (rapid…

Blits.ai offers (Alternative, Sentiment) — Graph-scale per deployment. The page's own worked example sizes a single mid-size retail bank at 2,000 reported scam-proceeds cases per year; across 6 deployments the labelled case and disposition layer is therefore plausibly tens of thousands of adjudicated mule cases, sitting on a continuously-updated node-and-edge graph built from full onboarding and payment data at each institution. Request further research for aggregated node, edge and disposition counts.

Training graph neural networks for first-party fraud and AML on real cross-bank graphs rather than synthetic ones, mule-network and layering-chain detection benchmarking, cross-institution fraud-graph link-prediction, reimbursement-liability modelling under UK authorised-push-payment split rules, scam-proceeds flow simulation and interdiction-timing models, investigator-triage automation, and evaluating whether a national or regional mule-data consortium would materially improve detection

The data is with 6 years of history.

Coverage spans Europe, US, UK, APAC; Application Software, Diversified Banks; alternative, sentiment; equities.