Data Management panel highlights
And now... for something somewhat different! Data Management.
A few months ago were assembled some of the thought leaders in Seattle (and one from Charlotsville VA) to discuss Data Privacy, Governance and Management -- I had the privilege of being their moderator. This section, the last in the series, focuses on Data Management. The panelists were (L to R in the video):
Michael Uschold: Enterprise Ontologist, Semantic Arts
Art Freas: Enterprise IT Architect, F5 Networks
Ron Hutchins: VP IT, University of Virginia
A selection of the key messages from the panel discussion:
Data, Storage and Metadata
The starting point: "Lots and lots of storage." How do you build on top of that the layers that allow you to understand the rest of those functions of applications, ownership, uses, privacy, security, reporting and governance.
The CMMI Data Management Maturity scale in a nutshell
"1. somebody somewhere in your organization has done something
2. an organization within your organization has done something
3. your whole organization does something
4. your organization measures what they do and
5. they actually use what they measure"
Most places to meet privacy regulations and best practices are at Level 3.
"I never approve data, I approve the use of data."
CI/CD (Continuous Integration and Continuous Delivery) just accelerates a problem you may or may not have... "The important things are simple, the simple things are hard."
"We have to really get not only the data democratized, you have to get the metadata democratized."
Data Silos and Knowledge Graphs
"The main thing is that the data is in silos everywhere."
Example: The Risk Department: they track different categories of risk, they have taxonomies for risk, for processes..." Risk data in silos, application for rules and issues, application for incidents, application for controls. Let us say you ask the question: "What kind of things are going wrong in this region, or what technology assets support this process that supports this other process in this region?" Knowledge Graphs serve as the metadata, not for one database, or for one application, but for all the applications and all the data.
"Rule #1- IT doesn't control access, business controls access." Rule #2, See Rule #1.
IT only really empowers the business to help you get to granular control. Otherwise, "data becomes sedimentary as in nature where things just settle out in the water and that never ends well..."
"Technology as it is today, is a driver for (data) silos"
On finding sponsors for Data and Graph projects
"Find someone who inherited the problem but didn't cause it, as your sponsor... because they're really invested, but they're not emotionally ashamed of it."
To summarize and paraphrase Ron Hutchins, a technology strategy on it's own doesn't make sense, technology should be there to support the business and operational strategies of the organization.
Links to the previous panels:
####