Why You Don’t Need A Data Mesh: How To Democratize Your Data With Modern Master Data Management
summary
Argues that modern Master Data Management already delivers data-mesh goals for most firms, based on three enterprise implementations (each 100k+ employees) that used Azure Data Factory, Azure Data Lake, Azure Purview for cataloging, CluedIn (graph-backed MDM) for automated linking and domain consolidation, and Synapse/Databricks for analytics. Reports rolling back parts of mesh designs, finds domains never fit a canonical model, orchestration (Kubernetes) always reintroduces centralization, and claims the extra 10% mesh complexity has poor ROI — one customer saved ~$1M by using access control instead of replicated infra.
AI-written summary · about
chapters (6)
- 0:00Introduction and contrarian view
- 4:13What we’re really trying to fix
- 9:07Modern MDM vs data mesh
- 14:27Azure stack demo walkthrough
- 19:53Six core lessons learned
- 24:31Q&A and closing
related talks
20mEmbracing Data Mesh: A Modern Strategy for Scalable Data Management
47mModern Data Architecture: Breaking the Big Data Monolith
23mHow To “Data Mesh”: What You Need To Know To Make It Real
36mData Mesh and Its Superpowers
50mData Mesh: the new paradigm for data management
48mData Mesh Paradigm Shift in Data Platform Architecture -
41mWhat is a Data Mesh - And How Not To Mesh it Up
51mData Mesh – Successful with People, Process, and Technology