Case Study: Zephyr AI achieves scalable HIPAA-compliant de-identification with Privacy Analytics Eclipse software

Zephyr AI creates 10 de-identified data iterations with Privacy Analytics

Zephyr AI, a high-growth company applying machine learning to health data, needed a scalable, self-service solution to de-identify protected health information (PHI) for its machine learning projects. The challenge was to avoid costly consulting fees and be able to perform multiple data transformations in-house to comply with regulations like HIPAA. They turned to Privacy Analytics for a solution.

Privacy Analytics implemented its Eclipse software module, which provided Zephyr AI with a flexible, self-service tool for de-identification. The solution allowed their team to run multiple experimental iterations on data sets to balance utility with privacy risk. As a result, Zephyr AI gained an on-demand ability to de-identify data, creating half a dozen data sets for model training without external service fees and preparing them for future partnerships.

Zephyr AI