The software arm of Capital One has expanded the availability of a tool that enables IT departments to improve their data security using data tokenisation.
The business-to-business software division of the financial services company has made its Databolt data tokenisation software available on two major data platforms – Databricks and Snowflake.
Capital One describes its implementation of data tokenisation as the process of replacing raw data with a digital representation. In a blog post discussing the benefits of the technology, the company said: “In data security, tokenisation replaces sensitive data with randomised, nonsensitive substitutes, called tokens, that have no traceable relationship back to the original data.”
Application areas include using data tokenisation to secure corporate data when training artificial intelligence (AI) models and protecting personally identifiable data, to comply with data protection regulations like General Data Protection Regulation and payment card regulations such as PCI DSS for ecommerce transactions.
Tokenisation is seen as an alternative to encryption, but is generally easier to integrate into existing IT systems. It effectively gives cyber security chiefs the ability to remove sensitive data from IT systems, which reduces the impact of data loss due to an IT security breach. It’s a technique used in financial services to protect payment data.
Among the benefits, according to Capital One, is that tokenisation preserves the length and format of data. Data tokenisation also maintains database relationships. This means it can be implemented in existing IT systems and applications without breaking how the applications process the data.
Analysis from McKinsey suggests that tokenised market capitalisation could reach around $2tn by 2030, driven by tokenisation of financial assets. Tokenisation is also a fundamental part of how large language models work where it is used to convert words and sentences into numerical values that can then be processed.
Capital One said its implementation of data tokenisation through Databolt enables companies to tokenise sensitive data directly within Databricks and Snowflake, making it easier for companies to protect their sensitive data where it resides. According to Capital One, this means IT security leaders can strengthen data security without slowing down innovation.
Desikan Madhvanur, senior vice-president, and chief product and technology officer at Capital One Software, said: “Today’s companies are managing data across a vast ecosystem. Integrating Databolt with Databricks and Snowflake is key to helping companies secure their data where it resides so they can confidently build applications and deploy AI models knowing their data is protected.”
Databolt provides data tokenisation via the Databricks Unity Catalog. Capital One said the integration allows Databolt customers to define tokenisation policies, ingest Databricks user groups for role-based access control, initiate tokenisation jobs and configure workflows.
It’s also available on the Snowflake Marketplace. Here, Databolt uses Snowpark Container Services and the Snowflake Native App Framework to provide native integration with the data platform. According to Capital One, the integration means sensitive data does not need to leave a customer’s Snowflake environment.
Capital One said the integration of Databolt on the Snowflake platform allows customers to deploy their tokenisation engine directly in their Snowflake environment, define access for tokenisation based on pre-existing Snowflake roles and access tokenisation functionality through user-defined functions.