Tests that work
Fictitious but valid IBANs and card numbers: format and check digit validations pass, with no real data.
Banking and insurance
Anonymize your financial data on your own servers
Customer files, transactions, claims files:
NymData detects US bank account and routing numbers, credit card numbers, IBANs and identities, then replaces them on your servers.
With realistic pseudonymization, the generated IBANs and card numbers pass format and check-digit validation.
Your test applications accept them.
Before: real banking data, in red
| Customer | IBAN | Card |
|---|---|---|
| Emily Brooks | FR76 3000 4000 3123 4567 8901 429 | 4970 1012 3456 7893 |
| John Carter | FR76 1820 6000 4199 0857 3412 035 | 5132 8800 4411 2097 |
Copied as is into a test database, the file exposes real accounts
After: fictitious, valid-format French IBANs and card numbers
| Customer | IBAN | Card |
|---|---|---|
| Laura Bennett | FR76 1027 8061 2100 0204 5110 182 | 4539 5781 2204 6631 |
| Ryan Mitchell | FR76 3006 6100 4100 0105 6630 209 | 5425 2334 3010 9903 |
Valid formats: application checks pass
- IBANs and cards detected
- Valid test data
- Same customer, same value
- On your own servers
How do you anonymize banking data for testing?
You need to replace IBANs, card numbers and identities with fictitious values that follow the expected formats, so your applications keep working.
With NymData, installed on your servers, realistic pseudonymization generates IBANs with correct check digits.
Card numbers are valid under the Luhn algorithm, with no fragment of the original.
The same original value always produces the same fictitious value.
The principle
Financial data gets copied everywhere
Application testing, vendors, auditors, reporting: account and transaction files travel far beyond production.
Every copy of real data is one more risk.
Consistency across files
The same original IBAN always produces the same fictitious IBAN: reconciliations across files and tables stay accurate.
Controlled sharing
Auditors, integrators and software vendors receive a copy with no identities and no bank details.
Documents covered
What finance and insurance teams anonymize
The formats that travel the most.
- Account and transaction files in CSV or Excel. Excel and CSV
- Customer and policy databases in PostgreSQL or MySQL, copied to a test database. Databases
- Claims files and letters in PDF or Word. PDF
- Presentations and reporting in PowerPoint. PowerPoint
Use cases
When should you anonymize this data?
Application testing
Populate test databases with valid, fictitious banking data.
Audit and compliance
Give auditors samples without customer identities.
Integrators and software vendors
Share realistic data for a migration or configuration project.
Analytics and models
Work on transaction histories without exposing customers.
Framework
Good to know
Generated IBANs are synthetic French IBANs with a correct French bank key (RIB key) and IBAN check digits; card numbers follow the Luhn algorithm.
They are meant for testing, with no fragment of the original values.
Pseudonymized data is still personal data under the GDPR: the systems that receive it must remain subject to your usual security controls.
Which method should you choose? · Pseudonymization vs anonymization: what is the difference?
Scope
What NymData handles for you
- SupportedIBANs and payment cards
- Detected automatically, along with US bank account and routing numbers, company registration numbers (such as the French SIRET) and financial identifiers.
- SupportedValid fictitious values
- IBANs with correct check digits and Luhn-valid card numbers, with no fragment of the original.
- SupportedDatabases and files
- CSV, Excel and PostgreSQL or MySQL databases copied to a destination database.
- SupportedOn your own servers
- No banking data is sent outside.
FAQ
Frequently asked questions
Are the generated IBANs valid?
Yes. With realistic pseudonymization, NymData generates French-format IBANs with a correct RIB key and IBAN check digits, with no fragment of the original IBAN.
Do the fictitious card numbers pass validation?
Yes. Generated numbers follow the Luhn algorithm that applications use to check a card number.
Does the same customer keep the same fictitious value?
Yes. Generation is deterministic within a project: the same original value always produces the same fictitious value, in every file and every table.
Which financial data is detected?
US bank account and routing numbers, IBANs, credit card numbers, company registration numbers such as the French SIRET and other financial identifiers, plus names, addresses, emails and phone numbers.
Is any data sent outside?
No. NymData is installed on your servers: no data is sent to NymData or to any cloud service.
See NymData prepare banking test data, live
A demo on sample files, remote or on site.