the semantic layer your AI needs
summary
OpenMetadata claims to become the AI semantic layer by shipping an MCP server that ships with the install and exposes JSON‑schema metadata, lineage and auth so LLMs can read it; a live demo connects Redshift, dbt and Superset, finds a broken dashboard root cause via column-level lineage, and rebuilds a corrected chart using a glossary term filter. Project stats: 11,000+ community members, 2,000+ orgs; contracts are beta and metric lineage isn’t yet auto extracted.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 1:03Open Metadata problems
- 3:09Open Metadata platform
- 4:15Open Metadata MCP server
- 6:27Demo setup
- 7:49Table metadata and lineage
- 9:15Debugging dashboard with AI
- 13:07AI visualizations
- 16:44Glossary terms for context
- 18:51Semantic layer for AI
- 19:35Closing and open source
- 20:21Q&A lineage and procedures
- 21:49Q&A MCP server and lineage
- 23:15Q&A cross-platform lineage
- 24:17Q&A manual lineage
- 25:21Q&A schema versioning
- 26:03Q&A metric lineage
- 27:07Q&A Metabase issues
related talks
32mIntelligent Data Platform : Powered by Semantic Layer, MCP & Agents
29mMetadata Systems for End-to-End Data & Machine Learning
24mMetadata driven framework for data discovery and data quality
37mMetaOps: Metadata Operations For End-To-End Data & Machine Learning Platforms
28mThe Structure and Interpretation of ML Metadata
45mSemantic Layer: Make Your Data Speak with AI Without Hallucinations
27mThe Sanity Check: Observability for AI Data Pipelines
32mTruly Open Lineage