Data Mesh
Enabling data mesh principles
What is data mesh?
A data mesh model is a domain-oriented decentralization for data >access.
Data mesh shifts the responsibility of data access from the data or technology teams, to the business teams.
Data mesh architecture empowers business teams to perform cross-domain data analysis independently. Teams no longer need to rely on data and technology teams who are overburdened and unable to keep up with demand to gain expertise in every business domain.
This approach often results in faster insights, increased productivity, and improved business outcomes.
Data mesh principles
Data mesh as a concept is based on four principles:
Domain ownership
Mandates domain teams to take responsibility for their data, allowing them to own both analytical and operational data.
Data as a product
Treats domain data as public APIs, and domain teams are responsible for providing high-quality data to satisfy the needs of other domains.
Self-serve data infrastructure
A dedicated team provides domain-agnostic functionality, tools, and systems to enable domain teams to seamlessly consume and create data products.
Federated data governance
Promotes standardization and interoperability of all data products through the whole data mesh, creating a data ecosystem with adherence to the organizational rules and industry regulations.
The resulting data mesh modelling fosters collaboration between different teams by allowing them to use other domains' data products. This simplifies data references and lookups. Additionally, using data from downstream domains enables teams to analyze changes.
By aggregating data from multiple domains, teams can build comprehensive reporting and create more innovative data products for the business.
How data collaboration enables data mesh
Supporting data mesh principles, data collaboration liberates and connects data in a federated network to facilitate controlled access.
The Cinchy Data Collaboration Platform liberates data from applications and links it in a network, where data is managed as a product and is shared, controlled, and reused - without the need for integration.
With data collaboration, enterprises can enable data mesh in the following ways:
Data products
Data is organized into domains as the governance body and can be structured by how the business unit uses and needs the data rather than how an application dictates.
Data ownership
Data and domains can be owned by different business units so domain experts can own and maintain their own data.
Data-layer access controls
Who has access to what data can be controlled at every layer by the data owners.
Data change workflow
Data owners can enable change approvals and track those changes throughout the process.
Data is organized by domains and control is directly on the data. It’s truly data access at your fingertips, with full governance and controls at the data level, not the application level.
In an altruistic vision of the future with a data collaboration model, whether you have 500 applications or 5 million applications the number of silos and integrations remains at 0. The reality is that you can get there iteratively project by project by replacing data silos with a mesh-based approach to the structure and access of the data.
Watch more on data collaboration
See data collaboration in action!
Connected data without the effort, time, and cost of traditional data integration.