What’s the deal with data assets, anyway?

There’s a lot I’m excited about in the latest Harbr release. But there’s one thing in particular that I want to talk about: data assets.

It’s more than a bit of jargon. It’s fundamental to how people actually use data.

I’ll first tell you what we mean by that. I’ll then illustrate how the concept of data assets maps to the platform in Harbr 5.0.

Data assets: the building blocks of working with data

To risk oversimplifying, data assets are the things that people who work with data use. Things like tables, notebooks, SQL queries, files — these are the building blocks of working with data.

Our roots in data commerce have put a major focus on enabling data producers to build and deliver data products. But the vast majority of users of any Harbr platform are data consumers. Indeed, a given Harbr platform may have hundreds or thousands of data consumers. These consumers are working with data at various stages in the data product lifecycle, and they are often working with data that isn’t necessarily part of a data product. They may be looking at anything from raw data that hasn’t yet been processed for a specific use case, to a mature asset that you’d expect to be part of a data product, complete with packaging, subscription plans, etc.

In Harbr 5.0, data products are composed of data assets. A data asset may be used across multiple data products without needing to be duplicated — or it might never be part of a data product.

From working closely with our customers and end users, we came to the insight that — generally speaking — data consumers will log onto the platform because they’ve been given access to some data product. Or they may want to use the platform to access or share a data asset. In fact, we’ve increasingly seen that consumers want — and need — to do both to unlock value from a product or asset. Therefore, we wanted to give users an instant view of the data assets and products they have access to.  

My collection gives a file explorer view of all the assets that you can work with individually, share with colleagues, or collaborate on. Here’s how it looks for a data consumer who wants to see the data assets available to them:

‍

My collection shows the data assets you have access to.

Looking a little closer, you’ll see that products and assets are both visible in your collection. If we expand a data product, we can then see the assets associated with that product. In this example, the data product ‘Median Sale Prices, United States’ contains a Python notebook and four tables.

‍

This data product contains several data assets, including a Python notebook and several tables.

Data assets: what and where?

It’s really important that the way Harbr handles data assets reflects the real world our users work in. In so many use cases, users need to work with data across boundaries — organizational, technological, or commercial. Therefore, we built Harbr 5.0 to be able to work with data assets wherever they are located.

A key aspect of data assets is that they can be located in various places:

• Cloud storage, such as AWS S3, GCS on Google Cloud Platform, or Azure Blob Storage
• Data warehouse or data lake, such as Snowflake
• On the user’s desktop — real or virtual
• As part of a Task, which are automations

You can also select a connector (which is dictated by the source) and the type of asset (table or file).

‍