Data Products – A Fireside Chat

So data intelligence to me is a combination of a platform that provides for discovery, for governance, for quality, for observability, and the ability to stage data products and to sell data products, um, which is a thing that we see as a, an emerging need in the market. What does data product actually mean in your view? Everybody talks about democratizing data, but the reality is that every department within an enterprise will have analysts that are crunching those numbers day in and day out.

And we haven't reached a point where the average business user can find and use the data assets that could be available to them. So a data product is essentially a data analytics asset that can be self-contained and that can be exposed to users in a way in which they can very easily make use of the contents.

So data products are, um, could be a data set, it could be a dashboard, it could be a report. Data products are governed by what we call a data contract. And a data contract lays out a description as to what it is.

It may have version information. It will tell you who the data owner is and who the business owner is for this particular asset. It'll then describe the terms under which that data product can be used.

Uh, it will tell you the quality. So maybe the financial service is a hundred percent accuracy a hundred percent of the time, but it may say, but this data's only good for 24 hours, 30 days, whatever it may be. And, um, the contract lays out also samples of what data might look like, um, and then some tests that you can make sure the data that you're receiving is, is accurate.

The industry has not done a good job with defining data products. We, uh, it, it's a pretty overloaded term. Um, many folks Would think of data sets when they think about data products.

Many people would think about data products as products that are somehow powered by data, but not necessarily the manifestation or, or of, of the data or the manifestation of insights that, that they're actually consuming. So it's good to see that you have a much broader view of what data product can be, and especially if it can be defined, if it's the right framework and if it has the right boundaries, then you can apply policies, uh, on, on those data products. You can actually have contracts.