


Enter PHEMI Systems—a company that simplifies the management of sensitive data for healthcare enterprises. Their trustworthy Health DataLab makes data-processing activities secure by controlling access to information based on the attributes of the accessor and ensuring data privacy and project governance. As such, data can be leveraged in multiple use cases without compromising privacy. More importantly, PHEMI manages compliance obligations like other data privacy solution providers while maintaining data-subject anonymity throughout data analysis, modeling and machine learning.
PHEMI uses a combination of data classification and meta-tagging to indicate data sensitivity and identify information that needs anonymization (and to what level). This plays a pivotal role in preventing unmanaged distribution of sensitive data, limiting the potential dangers of data leakage, while ensuring NSA level security for healthcare data. "The platform was developed by enterprise software architects who were visionaries in building big data platforms. They utilized their deep data expertise to design the Health DataLab a one-stop-shop for all healthcare research needs," says Rudy Potenzone, Chief Strategy Officer, PHEMI Systems.
PHEMI's Health DataLab is a cloud-based platform built with Privacy-by-Design principles on NSA developed technologies. It includes de-identification algorithms that hide restricted personal information, anonymize data and create dataset-specific or system-wide pseudonyms to mitigate the chances of leakage while sharing data.
As a result, analysts are equipped with the opportunity to leverage data without violating privacy guidelines. "Within the platform, we can ingest data, govern it, and provide access only to authorized users," Potenzone adds. "By combining best-of-breed open-source software, such as Accumulo, Spark, and Hadoop, our solution easily handles the huge influx of sensitive data, making it more scalable."![]()
Within the platform, we can ingest data, govern it, and provide access only to authorized users
More importantly, the platform incorporates cutting-edge AI algorithms to generate synthetic data and to identify PHI in unstructured reports. In spite of having the same characteristics as the real patient data, synthetic data is de-identified. " Health DataLab enables researchers to understand the dynamics of the patient population, while protecting patients' privacy," Potenzone mentions.
Today, several healthcare entities utilize Health DataLabin their research with success. To put things into perspective, Potenzone highlights their collaboration with a client where researchers wanted to accumulate data from over 200 diagnostic laboratories, analyze it, and make it accessible through dashboards for managing pandemic response. But, the system that the company was using did not allow them to securely accumulate information from such a variety of sources. PHEMI's platform allowed them to perform this complex task of pulling data within hours and helped make the data accessible to team members.
Propelled by such unwavering instances of client success, PHEMI is envisioning a promising future. The company will be improving the capabilities of the platform by incorporating advanced data science principles. "Healthcare data is the most ubiquitous data in the world. Our goal is to help clients leverage the immense potential of these data to initiate a new era of patient health and advances in medicine," Potenzone concludes.
Company
PHEMI Systems
Headquarters
Vancouver, BC
Management
Keith Elliston, CEO and Rudy Potenzone, Chief Strategy Officer
Description
The company simplifies the management of sensitive data for healthcare enterprises. Their trustworthy Health DataLab makes data-processing activities secure by controlling access to information based on the attributes of the accessor and ensuring data privacy and project governance. As such, data can be leveraged in multiple use cases without compromising privacy. More importantly, PHEMI manages compliance obligations like other data privacy solution providers while maintaining data-subject anonymity throughout data analysis, modeling and machine learning
