Term

Data Mesh

Data Mesh is a decentralised approach to analytical data, named by Zhamak Dehghani in 2019. It treats data as a product owned by the domain teams closest to it, replaces a central data team with federated computational governance, and standardises the wiring through a self-serve platform. Four principles: domain ownership, data as a product, self-serve infrastructure, federated computational governance. Take one out and you do not have a mesh.
Reviewed by 7wData

Why it matters

Data Mesh is the counter to the centralised lake or lakehouse pattern, where one platform team owns the pipelines and becomes the bottleneck for every data question. The pain is organisational, not technical: the central team cannot scale fast enough, so domains either wait or build shadow pipelines. Mesh pushes ownership back to the domains that understand the data. The load-bearing word is socio-technical. It is half org design, half platform engineering. Vendors selling “Data Mesh in a Box” are quietly selling the platform half and hoping you do the org half. You usually do not.

Where you’ll encounter it

Three contexts. A CDO pitches it as a reorg plan when a central data team is visibly overloaded. A vendor brands a catalogue or lakehouse “Data Mesh-ready”, which mostly means domain-scoped access and product metadata. An engineering team adopts it accidentally, treating their domain’s data API as a product with versioning, SLAs, a consumer contract. The common pitfall: teams pick up the tooling without the org design, end up with a fragmented data lake plus catalogue overhead, conclude Data Mesh does not work. The tooling was never the half doing the work.

Part of the 7wData AI Glossary. Tracking how concepts like this move in the expert conversation: daily signals at ins7ghts.com.