Case study · Financial services · Data security

Tokenization for PCI compliance at a leading financial firm

A leading financial firm tokenized Social Security numbers, card data, email addresses and names across Hadoop, Kafka and relational databases, using Protegrity, to reach PCI compliance.

Industry
Financial services
Capability
Data security, compliance
Platform
Protegrity, Hadoop, Kafka, RDBMS
Standard
PCI

Context

The client is a leading financial firm. It wanted to strengthen the security and privacy of sensitive customer data while maintaining Payment Card Industry (PCI) compliance, and engaged VerticalServe to implement a tokenization solution using Protegrity.

The problem

Sensitive data elements, including Social Security numbers, credit card information, email addresses and names, flowed through several storage and processing systems. The firm needed to protect them everywhere, keep the data usable in its protected form, and still let client-facing applications reach original values when necessary, with a record of every such access.

What was built

The solution has five parts.

Flow: SSNs, card data, emails and names pass through Protegrity format-preserving tokenization, policies and de-tokenization libraries, and are used in tokenized form in Hadoop, Kafka and relational databases, with every de-tokenization audited. SENSITIVE DATA SSN Card data Emails and names PROTEGRITY Format-preserving tokenization Policies and rules De-tokenization libraries DATA ECOSYSTEM Hadoop Kafka Relational databases Audit log of every de-tokenization · real-time alerts

Flow: SSNs, card data, emails and names pass through Protegrity format-preserving tokenization, policies and de-tokenization libraries, and are used in tokenized form in Hadoop, Kafka and relational databases, with every de-tokenization audited.

  • Tokenization with format-preserving masking: Protegrity’s tokenization protects sensitive data while keeping its format, so the data stays usable.
  • Policies and rules: tokenization policies configured for each data element: SSN, credit card information, email addresses and names.
  • De-tokenization libraries: libraries for client-facing applications to access original data when necessary, with encrypted transmission and strict access controls.
  • Audit logs: every de-tokenization is logged, with real-time alerts on suspicious activity or unauthorized access attempts, and tools for the firm to analyse and monitor the logs.
  • Use across the data ecosystem: tokenization integrated with Hadoop and Kafka data processing and configured for relational databases.

Stack

Protegrity Hadoop Kafka Relational databases

How it was delivered

The work started with assessment and planning: VerticalServe analysed the firm’s data architecture and storage systems, identified the sensitive data elements and how they flow through the organisation, and produced a detailed implementation plan and timeline.

Tokenization was then implemented and integrated with the existing data processing pipelines. The de-tokenization libraries followed, with thorough testing of both their function and their security.

Audit logging and alerting were put in place for all de-tokenization activity. Finally the solution was integrated with Hadoop, Kafka and the relational databases, and its performance and compatibility with the firm’s data ecosystem were validated.

The result

  • PCI compliance achieved, reducing the risk of penalties and reputational damage.
  • Stronger security and privacy for sensitive customer data through tokenization and format-preserving masking.
  • Secure, streamlined access to original data for client-facing applications through the de-tokenization libraries.
  • Better visibility and control over access to sensitive data through comprehensive audit logs.
  • Integration with the existing data ecosystem: Hadoop, Kafka and relational databases.

The original write-up reports these outcomes without figures.

What it means for a similar team

  • Map where sensitive data flows before choosing where to protect it. The assessment in this engagement set the scope for everything after.
  • Preserve format when you tokenize. It lets existing pipelines and schemas keep working on protected data.
  • Make de-tokenization the controlled exception: a library, an access rule, and a log entry every time.

This case study was first published on our Medium publication. Original article

Engage

Start a similar POD

Tell us the outcome you want and your constraints. We will reply with a discovery proposal.

Talk to us

Working on something similar?

An Innovation POD is a dedicated team that builds inside your environment. Tell us what you need to ship.

Start a POD