Bank transactions/Synthetic
Bank and card transactions
An agent-based simulation of US retail banking. Customers have a segment, an income schedule, and a home metro. Accounts are checking, savings, or credit card. Cards are an opaque token plus a random last four digits. No card numbers are generated. Payments run on card-present, card-not-present, ATM, ACH, wire, peer-to-peer, internal transfer, and cash deposit. Merchants carry an ISO 18245 category code. Approved amounts reconcile to each account's opening and closing balance. The standard specification is 10,000 customers over 180 days. Other sizes are built to order. Fraud and AML labels from the same run are the fraud scenarios set.
- Volume
- 10,000customers
- Formats
- CSV · Parquet · JSONL
- Label
- Synthetic
Purchase
Sold on Datarade. This site does not take payment.
Placeholder — listing URL not set
Schema
Column preview for the launch specification. Names and types describe the file. They are not a sample of rows.
- txn_idstring
Identifier for one transaction leg.
- ts_utctimestamp
When the transaction occurred, in UTC.
- account_idstring
Synthetic account. Not a real account.
- railstring
Payment rail, such as card-present, card-not-present, ACH, wire, or ATM.
- directionstring
Debit or credit on the account.
- amountdecimal
Positive amount in the account currency.
- currencystring
ISO currency code. USD in the current generator.
- auth_resultstring
Approved or declined. Declines are rows, not dropped.
Use
- Test ledger, payments, and account software
- Build features on synthetic card and bank activity
Generation
Synthetic customers, accounts, and payments. No real cardholder data and no real card numbers.
- Every dataset is 100% synthetic. It is not copied from real customers, cardholders, or patients.
- Each product is labeled as synthetic in the title, the description, and the file metadata.
- Delivery includes a generation note: the methods and algorithms used, the schema, and known limitations.