For Mark Clare, the data fabric he’s putting in place at Quest Diagnostics Inc. is the realization of a vision he has been pursuing for more than a decade.

Clare, a veteran data and analytics executive who joined the clinical laboratory early this year as enterprise head of data and analytics, is passionate about giving business users and customers self-service access to the data and insights they need. But that’s hard to do when data is scattered around the organization, as it was at Quest. 

This wasn’t a new challenge for Clare. Over the past decade, he has seen corporate data ecosystems explode in size, complexity and velocity. Enabling a business via an enterprise data warehouse or lake became prohibitive in terms of speed, talent and investment. Clare saw the need for a data fabric that could rapidly integrate a distributed digital ecosystem and enable business analytics quickly.

“Then the pandemic drove data officers to find ways to improve data and insights speed and agility further,” Clare said. “Post-pandemic, that business appetite has only increased, and we need new architectures like data fabric to meet this.”

Clare joined Quest just as generative artificial intelligence was breaking onto the scene. With diverse experience spanning healthcare, financial services and consulting, he has focused on helping the company become more data-driven while now also enabling AI adoption.

The new chief data officer found that data was housed across an ecosystem, with some on-premises and some in the cloud, managed by an assortment of software platforms. It was difficult to wrangle data rapidly into a form that would enable more insights to be discovered. But it was also a chance to create the distributed data fabric he had pursued in previous assignments. 

“You want to make data access simple, secure and consumable in a timely manner,” Clare said.

Seeds of opportunity

Gartner Inc. defines a data fabric as “an emerging data management design for attaining flexible, reusable and augmented data integration pipelines, services and semantics… across multiple deployment and orchestration platforms and processes.” It basically connects to data sources where they live and allows analysis to be conducted without the overhead of building data pipelines.

Quest Diagnostics’ Clare: “You want to make data access simple, secure and consumable.” Photo: LinkedIn

The job of building those pipelines falls heavily on data engineers, and traditional methods can choke efficiency. Clare draws an analogy to an iceberg. “What’s above the waterline is analytics and AI,” he said. “Below the waterline is all the data engineering,” plus all the necessary data governance, risk and compliance companies need.

Pipelines can take weeks or months to construct, but “business users want to measure delivery times in hours to days,” he said. “Do we want to build tens of thousands of pipelines as I have in the past, or deploy a fabric that has connectivity to all necessary endpoints?”

Clare turned to Promethium Inc., a developer of a data fabric, for help. Promethium says its data fabric provides a unified, consistent view and access to diverse data from across multiple sources and environments.

That creates a foundation for users to build data products that can access data from multiple sources quickly using generative AI. Its metadata-driven approach is centered on an augmented catalog that keeps track of all the data in the organization.

“I realized this is the unicorn I’ve been chasing for years,” Clare said.

Industry in transition

Quest is implementing its fabric in a healthcare industry that is rapidly transforming. The industry’s success during the pandemic in bringing COVID vaccines to market quickly dramatized the value of agility and speed. The culture of the entire industry is changing.

“What I saw in healthcare during the pandemic is that companies required immediate insights and the data capabilities necessary to rapidly support this,” Clare said.

Quest is letting business outcomes drive the evolution of its strategy, connecting data sources needed to support specific objectives. One of the benefits of a fabric is that it can be implemented opportunistically over time. “The construct is not to build it all at once, but to put a core in place with the necessary connectivity points to generate data products swiftly,” Clare said. 

The Promethium Data Fabric supports both virtualized and physically connected data products depending on performance needs. It’s flexible enough to accommodate new data types and sources as they emerge, an important consideration in a fast-changing business. 

“With quickly advancing data and AI technologies, if I designed just for today it could be obsolete before I fully deliver,” Clare said. “I need flexibility and extensibility in that core to adopt emerging technical capabilities as data products on the fabric and turn off legacy.” 

The approach also helps smooth the path to the cloud. “Over time, I can move more and more data onto the right cloud platform or platforms and turn off older architectures,” Clare said. “This increases capabilities to support digital modernization, while focused on business outcomes first.”

With generative AI rapidly maturing, Clare also sees the opportunity to broaden data access to a much larger audience. “People want to have a conversational approach to access the data they need to get answers,” he said. “When you think of the promise of generative AI, isn’t that really what businesses have been chasing with their data strategy for decades?”

Photo: Quest Diagnostics