Dear friends - Today, I’ll write about one of the most widely discussed topics - Data Mesh. This is a bit complex topic to understand if you are new to the data world. I’ll try to explain it in a simple way.
What is Data Mesh?
Data Mesh is an approach for building your data ecosystem to address your operational & analytics use cases. However, it is not just a technology but is more of a organizational change that impacts people, processes and technology.
How is it different from what we have been doing so far?
Traditionally, data analytics ecosystems are implemented by central data teams that work with various business units within an organisation and build the ecosystem as per business needs. This central team designs the architecture, builds the platform, manages it & creates access policies to share data with business teams. The central team is accountable for the platform and the data within the platform.
The Data Mesh approach is to build decentralised data teams instead of a central data team. Each business unit or domain manages its own data & creates the required insights & analytics. They can also share this data with other domains/business units, thus using their data as a “product”
Data Mesh Core Principles
Data Mesh has 4 main principles
Domain as data owners
Data is owned by its business/domain & not by a central data team. Each domain is accountable for its own data.
Data as Product
Each domain can share it’s own data with other domains and can also use data from other domains, thus leveraging their data as a product.
Self-Serve Data Platform
There will be a central platform team that can manage the infra & help/guide domain teams to manage & process their own data so that they have the self-servicing capability to generate insights & share data with other domains.
Federated Governance
Since every domain manages its own data, a federated(common) governance policy applies to each domain to maintain required standards & to ensure that each domain can easily discover & understand data products from other domains.
Summary
Data Mesh is a decentralised, domain-driven approach to processing & managing data. Instead of central data & analytic teams, each domain/business unit manages its own data & is accountable for its own data. Each domain can also share its data with other domains thus leveraging their data as a true data product.
Want to read more? Go through below useful links
https://www.datamesh-architecture.com/
Hope you have liked this post, thanks for reading!
Nice Explanation... Thanks