Follow Datanami:
September 22, 2015

Phemi’s Big Data Approach: Never Trust, Always Verify

Alison Diana

To address organizations’ security, privacy, and data silo challenges, Phemi today unveiled its Zero Trust Data approach to big data which is based on the Zero Trust Networking model that demands adopters “never trust, always verify.”

Phemi’s Zero Trust Data model embeds and enforces consent, data sharing pacts, and privacy policies at the data level, according to the developer. Therefore, the solution encodes privacy and governance policies directly onto data as an organization collects it, and an enterprise can opt to encrypt data and make it inaccessible by default.

“The big data movement has opened up a lot of opportunity, but it has also introduced new security and compliance risks to the point where access and security has trumped the value of data integration,” says Stewart Bond, director of Data Integration Software Research at IDC, in a statement. “Phemi’s Zero Trust Data approach extends the zero trust networking concept to data where nothing is trusted, and everything is verified, bringing a new focus on data trust, availability, security and compliance in today’s big data environments.”

If users do not have the correct credentials they cannot garner information. Users’ metadata – such as their name, roles, responsibilities, and hours – automatically confirms or denies admittance to specific data via the application, said CEO Paul Terry in an interview.

“We can do zero trust and decide on a person-by-person basis or piece-of-data by piece-of-data basis who sees what,” he said.

Likewise the data itself is surrounded by metadata, particularly important to data-centric customers such as finance, health, and life sciences, said Terry.

“The competitive advantage is fundamentally how they find patterns in their data and that is the essence of big data,” he said.

Phemi ties together the supercomputing strengths of founder Terry – whose OctigaBay was acquired by Cray – with the expertise of a cardiologist to design a big data platform that can handle thousands of diverse data repositories, ranging from electronic medical records (EMRs) and SQL databases to x-rays, phone records, videos, and genomes, said Terry.

The platform, which can be deployed on any cloud, puts all incoming data into a library of big data, which it then secures. The system then catalogs the library, allowing authorized users to search petabytes of information in milliseconds, he said.

“This is essentially a small supercomputer. We can have processors computing the information in our data library,” added Terry. “We can create algorithms and deploy in situ.”

Today, about 60 percent of the 30-month-old company’s customers are healthcare organizations, he said. And while the Canadian business focuses primarily on the United States, it’s seeing growing interest in Europe where organizations must balance the “right to be forgotten” and other privacy rules against their desire to leverage insight from big data, said Terry.

But it’s healthcare organizations that exemplify Phemi’s primary focus and today the company announced it is working with the Personalized Medicine Initiative (PMI), which is based in the Life Sciences Institute of the University of British Columbia, to collect, store, and manage genomic and clinical data for the Pathfinder Project by Molecular You Corp. (MYCo).

MYCo is designed to provide 20,000 people with a molecular understanding of themselves over the next five years, while simultaneously giving scientists insight into disease detection and early intervention. MYCo is using Phemi’s Central Big Data Warehouse in combination with Hortonworks Data Platform (HDP) to collect and manage this array of information.

“This is an exciting time for medicine. With new technologies that affordably and quickly sequence a person’s genome, as well as big data technologies that enable us to make the most of our data, we are in the position to significantly improve individual patient outcomes, save millions in healthcare costs, and open the doors for new discovery, prevention and treatment methods,” said Rob Fraser, chief operating officer at PMI and co-founder of MYCo. “[Phemi provides] a secure big data solution that lets us easily collect and share data with the various research organizations contributing to this project and, at the same time, protect the privacy of our program participants.”

In addition to using Phemi’s Zero Trust Data approach for governance and privacy, MYCo plans to use the platform’s de-identification capabilities, according to the company. Participants, too, can see their own data, Terry said.

Phemi Central, which is built atop Hortonworks’ Open Enterprise Hadoop-based HDP, melds together full-life cycle data management and access control with HDP’s big data capabilities.

Datanami