Free reference · Data Lake + Knowledge Graph

The lakehouse, in detail

Sources, connections in, bronze, silver and gold, the catalog, and what reads from it. Zoom in, or print a copy for your team.

Download SVG
The lakehouse, in detail. Sources on the left: streaming (app and web events, devices, web analytics) and batch (production databases, ERP, CRM, billing, file shares) are internal; third-party SaaS apps, partner feeds and vendor APIs are external. Connections in: streaming ingestion, change data capture on databases, read-only snapshots of file shares, and API loaders; your production data never moves. Inside the lakehouse, in your cloud tenant: bronze keeps raw data as it arrived, read only and replayable. Silver builds a single customer view, big transaction tables, flexible schemas and history with slowly changing dimensions. Gold is cleansed, with common definitions and interfaces built for speed and accuracy. A knowledge graph links entities and documents, and Unity Catalog puts owners, access rules, lineage and the audit log on every table. Connections out: ML models trained and served from gold, BI dashboards, AI agents through purpose-built functions, your apps, and data sharing. Underneath: security as a layer, and standard patterns for code, tests, quality, monitoring and cost. SOURCES Streaming INTERNAL App and web events Devices and sensors Web analytics Batch INTERNAL Production databases ERP, CRM, billing File shares Third party EXTERNAL SaaS apps Partner feeds Vendor APIs CONNECTIONS IN Streaming ingestion events land in seconds Change data capture AUTO CDC on databases Snapshots, read only file shares, not copied APIs and loaders scheduled pulls YOUR PRODUCTION DATA NEVER MOVES Lakehouse YOUR CLOUD TENANT · OPEN FORMATS Bronze Raw, as it arrived Read only Replayable Change data capture by default Silver Single customer view every record merged Big, fat tables transactions merged Flexible schemas any type of data History slowly changing dims Gold Cleansed Common definitions Interfaces built for speed and accuracy Knowledge graph entities matched in silver, documents attached Unity Catalog owners · access rules · lineage · audit log, on every table CONNECTIONS OUT ML models trained and served on gold BI dashboards Power BI, Excel, Databricks AI agents purpose-built functions Your apps fast reads for products Data sharing partners, no copies Security as a layer: access granted on purpose, table by table and row by row; agents follow the same rules as people Standard patterns: code in your repo, tests and CI/CD, quality rules, monitoring, cost reporting

DataMind Technology · datamind.technology · Book a free discovery session: datamind.technology/book-discovery-session/

← Data Lake + Knowledge Graph · Free Discovery Session