Data Mesh in Practice

Part I. What Is Data Mesh and Why Do We Need It?

For at least three decades now, companies have been trying to get to a place where they can easily use analytical data to improve their business. For example, they try to utilize data about the behavior of their customers and about the usage of their products to generate clear and actionable insights that help them to run a better business and build better products. Although there are a few success stories in this regard, when talking to leaders and practitioners within the industry, we are faced with countless struggles that seem to go nowhere and create a lot of frustration. The now overused analogy of oil and data has triggered a rush of investment in big data technologies and business intelligence capabilities that often do not meet the expectations that were placed on them.¹

Data mesh is a fairly new approach to data architecture that decentralizes ownership of domain data while applying product think­ing to analytics data. It is mainly a solution to problems with orga­nizational scaling. The reason why data mesh is not focusing on addressing problems with technical scaling is that many of those challenges have already been addressed by the technological innovations of the past years, most notably by public cloud platforms and by parallel computation frameworks.

The following two chapters sketch the status quo of data archi­tecture and introduce the main concepts of the new data mesh paradigm.

¹ NewVantage Partners, Big Data ...