A Data Mesh Reference Implementation on AWS
summary
A concrete AWS reference implementation claims to make Data Mesh repeatable by offering one-click creation of data products from a declarative data product spec. It spins up an S3 bucket, Managed Airflow, EMR for Spark, a GitLab repo, registers the spec in Apache Atlas, and generates lineage; deployments take ~20 minutes. It includes a sidecar for field/row access control, experiments with key‑management encryption, and uses Open Policy Agent for governance; the demo works end‑to‑end but notes cost and tooling tradeoffs.
AI-written summary · about
chapters (8)
- 0:00Data Mesh on AWS intro
- 5:22Core principles
- 11:03Data product design
- 16:15Self-service platform
- 29:42Lifecycle and access control
- 35:22Future challenges
- 40:50Real-world cases
- 45:20Q&A
