Putting Theory Into Practice: Getting Started With Master Data In A Data Mesh
summary
Argues that mastering must come before a data mesh and shows a working pattern: a persistent ID service called by domain APIs that links divergent records (Salesforce, Jira) into one reusable master, plus domain-local mastering for sensitive KYC lists. Uses human-guided machine learning, pretrained match models, reference data and rules packaged as data products, orchestrated with pipelines (Python/Airflow) and stewardship feedback loops. Runs in the cloud to get on‑demand compute and elastic scaling; presented as a repeatable, deployed approach.
AI-written summary · about
chapters (8)
- 0:00Introduction and audience poll
- 0:43Data mesh recap and context
- 3:53Key logical entities and data products
- 10:00Where to master data in data mesh
- 16:40Central vs domain-driven mastering
- 19:28DataOps and skill sets for domains
- 22:39Cloud benefits for data mastering
- 25:08Key takeaways and wrap-up
related talks
23mHow To “Data Mesh”: What You Need To Know To Make It Real
51mData Mesh – Successful with People, Process, and Technology
52mFrom Data Mess to Data Mesh: Navigating People, Process & Platforms
34mBest and Worst Days of Data Mesh
43mUntangling Data Mesh: A paradigm shift in Big Data
20mEmbracing Data Mesh: A Modern Strategy for Scalable Data Management
38mData Mesh, Now
36mData Mesh and Its Superpowers