top of page

Data Mesh: A Renaissance, Not a Revolution



The buzz around 'data mesh' has grown louder in recent years, with many in the industry hailing it as a revolutionary new concept. But let's set the record straight: the principles underpinning data mesh are not as new as they seem.

 

For over ten years, I have been at the forefront of crafting data architectures that capture the quintessential principles of what is contemporarily regarded as a "data mesh". My pioneering approach has manifested through a variety of monikers—ranging from data virtualization to data fabrics, data hubs, and direct query architectures. During my tenure at a nascent venture that Deloitte absorbed in 2017, I introduced the concept of "Lean Data Architecture", which garnered recognition within the sector as "CBIG Lean". The core tenet has consistently been to streamline the organization and retrieval of data, making it instinctive and centered around business utility, enabling egalitarian data access with agility, while concurrently reducing data transit and simplifying the overall data architecture framework.

 

The notion of infusing object-oriented programming paradigms into data handling, leveraging metadata to craft reusable and business-aligned objects, is a tale as old as time (in technology years). We witnessed it during the era of expensive data warehouse appliances, where such strategies were critical to maximizing investments. Pioneers like Bill Inmon, Ralph Kimball, Doug Cutting, and Mike Caferella led the charge, addressing similar challenges with ingenuity and foresight.

 

What we're witnessing now is the narrative's natural evolution, driven by the relentless migration to the cloud and the decentralization of decision-making to the network's edge. The demand for methodologies encapsulated by the data mesh concept is on the rise. However, we must tread carefully. In our eagerness to sprint towards a gleaming future, we risk overlooking the value of our existing assets. The focus should be on enhancement, not replacement.

 

So, what is a data mesh? There are numerous interpretations, but I resonate most with the idea that it's a business domain-centric approach that enriches your engagement with data. Imagine a Socratic dialogue with your datasets—comprehensive, secure, and governed.

 

As we venture further into this domain, I remain excited to contribute to the evolution of this narrative. There's a collective goal here: to refine our capability, technology, and methodologies, ensuring they mature in a way that benefits the industry at large.

 

Let's continue the conversation, not as if we're embarking on a brand-new journey, but as if we're refining the path laid down by the giants before us.

 

Comments


bottom of page