On-Demand Webinar
How to Implement Data De-identification for the Modern Analytics Platform
We are only just beginning to see the potential benefits of machine learning and advanced analytics: rapid fraud detection, hyper-targeted personalization, predictive analytics for inventory planning and price optimization, breakthroughs in drug discovery and precision medicine. All of it is powered by the massive amounts of data being created every day by people, processes, and IoT devices - and much of that data can be considered “sensitive” or “Personally Identifiable Information (PII)” under various data privacy and industry-specific regulations.
Data platform architects must figure out how to make the organization's data available while still meeting compliance requirements. Attempted solutions usually involve complex ETL or static data masking; however, these solutions don't address the fact that completely removing the sensitive values also removes a lot of the data's utility.
The answer lies with data de-identification, the process by which data values are transformed or obfuscated in order to protect sensitive information while leaving it in a form that is still useful for the business. In this webinar, hosted by Lars George (Principal Solutions Architect for Okera), we'll cover:
- An overview of why traditional methods to redact sensitive data are unsuitable for the modern enterprise analytics platform
- Anonymization vs. pseudonymization, and when it’s most appropriate to use either one
- The case for dynamic policy enforcement of de-identification types
- Examples of specific de-identification policies being used by F500 enterprises, and how to author them in Okera’s visual policy builder
