Overview

Mutua Madrileña, also known as Mutua, is Spain’s largest insurance company, providing auto, life, health, homeowner, and accident insurance — as well as investment products — to over 5 million customers. Founded in 1930, Mutua has many times been named the best company to work for in Spain and continually scores top marks for the best corporate governance practices.

Problem 

Operating in the highly regulated insurance industry, Mutua is required to comply with stringent consumer privacy and data protection regulations. In particular, the company must mask all personal data used in non-production software environments such as development, testing, staging, or training. Masking all this data is an enormous undertaking since Mutua has over 500 million records in its SQL databases, and hundreds of millions more in legacy DB2 databases.

Solution 

Mutua searched for a data-masking tool that could scale to hundreds of millions of records across multiple database types. During the product search, the IT team realized a synthetic data solution would be more flexible and scalable than a simple masking tool. They saw that, unlike masked data, synthetic data could be used to build predictive models that maintained the original statistical distributions. An example of this would be using zip-code data mapped to policy types to extrapolate which insurance products to upsell in specific neighborhoods.

Mutua selected Howso because it is the most powerful synthetic data tool on the market. They also appreciated how easy Howso is to use for data scientists, test engineers, and other tech staff working on software projects.

Impact 

In the 15 months since Mutua rolled out Howso, the company has already created over 300 million synthetic records, and are well on their way to creating more than 520 million in the next few months. Mutua uses Howso across all of its multiple types of databases. Instead of just masking their data, Mutua’s tech teams create synthetic datasets with the same statistical properties as the original ones — except with no personally-identifiable information. This ensures all software projects use only fully compliant data. Already, Mutua has used Howso to avoid potential fines for non-compliance.

Apart from using synthetic data to meet regulatory compliance, Mutua also is exploring ways to analyze the data to grow their business. For example, they can create predictive models to forecast which products and services would be most beneficial to customers.