Plaid

Plaid — Consented bank-connected payment, identity and risk dataset generated as a byproduct of powering account linking for consumer apps, specialised here to gaming and prediction-market operators: per-deposit ACH transaction records with counterparty and funding-source detail, identity-verification (KYC/AML) records, and a derived risk layer of 80+ predictive attributes scored in real time before funds move. Continuous monitoring emits behavioural signals across three joined surfaces — identity, device and linked financial accounts — that detect fraud, bonus abuse and multi-accounting. As …

Plaid offers (Alternative, Reference, Sentiment) — Very large but undisclosed: an identity, device and linked-account graph spanning tens of millions of consumer connections, with continuous transaction and deposit event streams and 80+ attributes scored per event — order 10^9-10^10 scored events per year at a ~$575M revenue scale, with a correspondingly large cross-referenced actor graph. Plaid publishes no volume figures..

Fraud and abuse model training with verified money-movement outcomes (rare, since most fraud datasets are label-noisy), graph-based identity-resolution and synthetic-identity research, responsible-gambling and affordability assessment models using deposit cadence against income signals, AML and suspicious-activity monitoring models, consumer cash-flow and spending-velocity indices as a macro alternative-data signal, deposit-decline and payment-cost optimisation research, and prediction-market participant-behaviour studies. For a research buyer the differentiated angle is affordability scoring — labelling whether a deposit is proportionate to observed income — which underpins emerging duty-of-care regulation.

Coverage spans US; Transaction and Payment Processing Services, Financial Exchanges and Data; alternative, reference, sentiment; equities, fixed_income.

The supplier's own access route is a contact form. Request a sample through Brickroad to start.