Valorise data with Data Mesh 2/4
In the previous article, we explained the principle of " Domain specific knowledge''. This article will explore the principle of "Data considered as a product"
Make data understandable and accessible to anyone through data product
Once a squad is created, as discussed in the previous article, its primary objective should be to elaborate data products that are easily discoverable, understandable, accessible, and usable. A data product is typically the result of a set of pipelines extracting, transforming raw data, and exposing them and their metadata. By doing so, the data product can be seen as an off-the-shelf product accompanied by KPIs and SLAs, the squad being responsible for those. Data product approach helps gaining time and, with a sense of ownership higher, the quality of data is better. Moreover, the exposition of data product through API enables an easier access and search.
The Marketing squad, after fetching the data they need, will give access to pre-aggregated data and information about the transformation they used to the rest of the company. Not only the Customer Communication Service (CCS) will get data they understand by themselves, but this gives the opportunity to other teams in the organization like the billing team to use them as well.
Now that we have accessible and understandable data, we need to ensure it stays that way, that the data can be used across the entire organisation and secured by defining a data governance.
[“Federated computational governance'' will be discussed in the next article]