Easy Metadata Management (Git for Data) Using Neo4j
summary
Built a metadata system on Neo4j Community Edition that captures table/column lineage, ownership, and schema types instead of relying on Bloom; it generates saved Cypher views and interactive hierarchies from bank→department→service→database→table→column graphs and answers “who owns X” and “where does this column come from” queries in seconds instead of the prior month of manual discovery. The implementation enforces type-safe identifiers and type-safe relationships, supports enums and dates, and includes a custom visualization/dashboard — it works live on production data.
AI-written summary · about
chapters (8)
- 0:00Introduction to metadata
- 1:33Choosing Neo4j
- 2:58High-level system design
- 3:42Data lineage and hierarchy
- 5:41Demo: Visualizing bank data
- 8:32Type-safe entities and links
- 11:02Demo: Live data operations
- 13:06Management and conclusion
related talks
46mJourney of metadata
24mMetadata driven framework for data discovery and data quality
52mWhat the Graph
16mA Data Management Layer for Neo4j
29mBreaking Silos: Using Neo4j to Unify Tax and Finance Data
35mUsing Semantic Metadata to Create an Automated Microservice Data Mesh
23mDisrupting Customer Data Platforms: Neo4j vs. the Status Quo
15mGraphRAG-Powered AI Agents: Transforming Corporate Metadata into Intelligent Knowledge