Stop Mummifying Data
Rather than worshipping data as sacred, exploit it!
Authors and Coauthors Listed Below.
Pyramids, castles, shrines, shopping malls, stadiums, museums — humans have always made monuments to the things they value.
Even as what we value has evolved over thousands of years, our primitive instinct to build around them remains. As we move deeper into the current millennium, we have a new “thing” that we value: data.
Following our primitive instincts, we built data warehouses, marts, lakes, cubes, and bricks. So entranced by its value, we fortified in every way we knew possible. To protect and pay homage to these data monuments, we built moats and vaults around them by replicating the function-based silos that dominate healthcare and other industries.
When you think of data lakes, you might think of an advanced synthetic topography. But they and their brethren are painfully old-fashioned — more like a landfill.
Designed around function-based divisions, contemporary data platforms are stuck in the 1990s. Storage, modeling, and normalization specialists mimic the traditional functional silos, creating bottlenecks for today’s as-a-service platforms. Bloated investments such as multiple warehouses, on-premises data marts, cloud-based data lakes, and disparate business intelligence tools only exaggerate this costly effect. Witness this trend across all industries: Business lines within industry verticals tend to have different solutions, perpetuating specialists’ need to feed data and technology platforms.
The result?
- A perpetual backlog of “reporting” needs, while the expensive enterprise data, analytics, and intelligence teams are buried in operations overhead and the forensics of unearthing the nuances of data requests.
- Layers upon layers of expensive data infrastructure
- Inability to operationalize and realize values on AI predictions. Without the associated action or intervention, prediction is just a mathematical exercise.
- Investment fatigue in full-stack technology solution for every problem
- High-value human assets used for data collection to fulfill the future promise of value
Caught by this evolutionary drive to fortify, we think we can do it better if only we do it bigger. Build, build, build.
What if data is not a thing that needs fortification?
Imagine a data system that mimics the human brain. The brain orchestrates multiple sensory feeds to generate precise algorithmic responses to the muscular and skeletal systems. This analogy allows us to think about data as a stream — a stream of enterprise consciousness. While downstream computation for trending and analysis is valuable, organizations can positively apply upstream cognitive computing to impact actions and interventions positively.
Exploit data for its value, right from its inception. Despite the need for historical analysis and perspective, data ingested and stored for analysis after the fact has already lost most of its value. By applying cognitive computing near the source and converting it from information into actionable intelligence, data becomes meaningful and perishable. Decentralizing data, so informed decisions can be made at the edge in real-time, activates “Smart Operations” allowing organizations to enable enterprise cognitive capabilities focused on critical operational outcomes.
We recommend strategically disabling the traditional silos in functional specialization, both from a technology and a talent perspective. This new philosophy will do more than automate and optimize workflows; it will prime them to become responsive and intelligent. Organizations that have spent the last decade warehousing data will require new and distinctive cross-functional, collaborative and cognitive data models. In the cognitive enterprise, data is recognized as a stream organized around outcomes.
The result:
- Upstream distribution and consumption of analytics to improve the quality of report/data requests. Data requests that are more directed, outcome-driven, business-relevant, and conducive to self-service.
- Decluttering (and decommissioning) layers of redundant technologies.
- Active integration to operational outcomes.
- Innovation in the algorithmic level.
Now, beyond all this, imagine having the capability to create adaptive smart workflows that morph to accommodate real-time changes in the environment. That is the promise of Cognitive Data Streams and Smart Operations.
Rather than worshipping data as sacred, exploit its value. Data should not be an abstract meme for value and should be decoupled from the memetics of the pyramid and Non-Fungible Tokens (NFT).
Let’s stop mummifying data in technology tombs and start exploiting it to make an impact in day-to-day life.
Author: Balaji Ramadoss, Ph.D. Founder and CEO of Edgility Inc.
Coauthors:
Peter J. Pronovost MD, PhD, FCCM, MacArthur Fellow, Chief Transformation Office at University Hospitals, Cleveland and co-author of Safe Patients, Smart Hospitals.
Justin Falk, is the Chief Technology Officer at Edgility InC.
Edmondo Robinson, MD, MBA, MS, FACP, Senior Vice President and Chief Digital Innovation Officer for Moffitt Cancer Center.
Beth Lindsay-Wood, MBA, CHCIO, is the Vice President and Chief Information Officer at Moffitt Cancer Center.
Jason Roos, MBA, is the Vice President and Chief Information Officer at King Abdullah University of Science and Research.
Sam Brown, MBA, is the Vice President, Logistics and Systems Operations at University Hospitals
Sarah Mihalik is the Executive Director, Healthcare Analytics & Innovation at University Hospitals
Wilfrido A. Moreno, Ph.D., is a Professor for the College of Engineering at the University of South Florida with a speciality in Systems Engineering, Industrial Controls, Digital Signal Processing and Reconfigurable Architectures.
