Secure data analytics and AI with Databolt for Databricks

In today’s data-driven world, organizations are constantly challenged to unlock insights from sensitive data while meeting privacy, compliance and security requirements. At Capital One, the challenges led us to invest in tokenization to secure our data while maintaining usability. Furthermore, the successful adoption and utilization of tokenization became the motivation of our latest product, Capital One Databolt. We didn’t stop with just building Databolt. We understood that there is no one size fits all tokenization solution. So we have partnered with Databricks to offer companies seamless, enterprise-grade tokenization to the Databricks Lakehouse Platform with Databolt.

Enterprises want to get value out of their ever increasing data volumes, but they also want to keep it secure. Tokenization can allow enterprises to de-identify sensitive information: such as PII, PHI, PCI and other sensitive data—while maintaining the usability required for machine learning, analytics and AI. Our tokenization product, Databolt, integrates directly with Databricks Lakehouse to allow users to seamlessly tokenize their sensitive data so it’s ready for AI and analytics workflows.

• Seamless Databricks integration: Databolt integrates with Databricks Unity Catalog, Databricks Workflows and Databricks supported User Defined functions (UDFs) to provide an end-to-end tokenization solution.

• Flexible deployment architecture: Easily configure Databolt using our customer portal while keeping your sensitive data and processing in your environment.

• Multi cluster support: Databolt provides the ability to work with Databricks provided compute to run tokenization workloads.

• Secure data analytics and AI-ready data: Databolt tokens maintain referential integrity, maintaining utility of tokenized data for analytics, machine learning and AI without exposing sensitive information to data consumers, data users and data scientists.

• Containerized application: Databolt is designed to scale both vertically and horizontally to take advantage of cloud technologies like Kubernetes.

If you’re already using Databricks, adding our tokenization layer is fast and simple. You can use:

• A helm-chart installer to package and deploy the tokenization service in a Kubernetes cluster

• A front-end to configure your sensitive fields, access policies, access controls and workflows

• A Databricks tokenization workflow package

• A Databricks UDF integration to define access to allow for detokenization of data

• A connector to ingest metadata from the Databricks Unity Catalog

• A secure, audited API or SDK to call tokenization functions directly

As businesses adopt AI at scale, the ability to secure data without compromising utility becomes critical. Together with Databricks, we’re making that a reality. To learn more or schedule a demo, visit our product page.