Among all the topics discussed at the Gartner Data & Analytics summit, one undercurrent caught my attention. It came up at least twice, in very different talks.
The first occurrence was in a presentation by an e-commerce company which made all of its data open to all employees and encouraged them to create innovative analyses of that data. The introduction of the General Data Protection Regulation (GDPR) last year has caused the company to restructure its data platform with stricter access policies, as they can't reasonably make the personal information of all their customers available to everyone in the organisation.
The other time was in a talk by a Gartner analyst, about different data architectures (e.g. data warehouses, data lakes, & data hubs). In an aside, the speaker remarked that, in the past, IT departments have tended to collects all their data together first and worry about governance later. He said that design of a data hub should instead start with data governance - i.e. who is allowed to access what, and how the data will be kept accurate and up to date.
It struck me that our University may actually be ahead of the game here. We have put governance at the heart of our data architecture. We have defined a data steward role and assigned data stewards to our core administrative data sets. We have a data governance steering group in place. We assign a confidentiality classification to all of our administrative data, and of course the implementation of the GDPR has given us a catalogue of the our personal information that we hold.
These are only the first steps. We undoubtedly have lots left to learn and do, in data governance as well as other aspects of the architecture, but at least in this respect we seem to have started out on the right path.
The first occurrence was in a presentation by an e-commerce company which made all of its data open to all employees and encouraged them to create innovative analyses of that data. The introduction of the General Data Protection Regulation (GDPR) last year has caused the company to restructure its data platform with stricter access policies, as they can't reasonably make the personal information of all their customers available to everyone in the organisation.
The other time was in a talk by a Gartner analyst, about different data architectures (e.g. data warehouses, data lakes, & data hubs). In an aside, the speaker remarked that, in the past, IT departments have tended to collects all their data together first and worry about governance later. He said that design of a data hub should instead start with data governance - i.e. who is allowed to access what, and how the data will be kept accurate and up to date.
It struck me that our University may actually be ahead of the game here. We have put governance at the heart of our data architecture. We have defined a data steward role and assigned data stewards to our core administrative data sets. We have a data governance steering group in place. We assign a confidentiality classification to all of our administrative data, and of course the implementation of the GDPR has given us a catalogue of the our personal information that we hold.
These are only the first steps. We undoubtedly have lots left to learn and do, in data governance as well as other aspects of the architecture, but at least in this respect we seem to have started out on the right path.
Comments