‘Distributed Stewardship’ Seen as Path to Data Democratization

Okera made its mark on the big data map by providing a universal access layer that gives administrators a single place to control what data users can access, no matter what tools or data sources they’re using. The company bolstered its data security story today with the introduction of distributed stewardship, which pushes the responsibility of defining high-level data access rules out to experts who know the data best.

There is a natural tension between centralized control and distributed access in the big data business. On the one hand, companies that remove barriers to data access and get the right data into the right set of hands can do very innovative things with data and gain a competitive advantage. This is the bright side of the “data democratization” process that so many companies talk about.

But on the other hand, regulations increasingly demand that companies place limits on data access in order to protect the integrity of that data, as well as to protect people’s privacy. Laws like GPDR, CCPA, and Virginia’s new Consumer Data Protection Act (CDPA) to impose fines and–perhaps worse–public humiliation of their brands when regulations are violated.

Sitting in the middle of that push-and-pull is Okera, which emerged three years ago to streamline that data access while simultaneously providing the necessary security controls. Simplifying the process of controlling who gets access to what data, no matter where it resides, is the number one goal of Okera, says Nick Halsey, who joined the company as its CEO about a year ago.

“If the policy towards the data changes, you want to be able to change it in one place and then have it be enforced consistently across all data source and applications, whether you’re using a BI tool on a relational database or a machine learning tool on object store, or you’re in the cloud or on prem,” Halsey tells Datanami. “None of that should matter. You should be able to…define the policy once and be able to have it then executed consistently across all the different use cases.”

With today’s introduction of distributed stewardship, Okera is putting a plot twist on that storyline of centralized control. While chief data officers (CDOs) and other executives will still have blanket control for some types of data access requests–such as declaring that card numbers and Social Security numbers will never be in the clear–access to other types of data will be handled by the folks who are experts in that particular data domain.

For example, if you want to have HR data made available in the organization for people to do analytics upon it, you need to be able delegate management of the access and the privacy of that data to the HR team, Halsey says.

“Or if you’re going to let the clinical researchers in a pharmaceutical company do machine learning on the clinical trial data, you better let the oncology research team manage the cancer research data, because IT doesn’t really know what are the characteristics of that data,” he says.

It’s about trusting your various team leaders to define intelligent data access policies that match the nuance of a particular data set, while simultaneously verifying that they’re doing things correctly, Halsey says.

“It allows you to delegate control over sub-groups of data and subsets of the data out into the organization, and then centrally report on it and be able to monitor who’s doing what with the data,” he says. “You can distribute responsibility, but centralize governance.