Information Integration is the discipline of determining which Concept each data record belongs to - and that depends entirely on Definitions. Part 1 of a three-part series on Definitions as the foundation of every Data Warehouse.
Subtraction at the operational layer, deepening at the modeling layer. The local Postgres sidecar is gone, deploy and execute are ~2× faster, --full-refresh now works at workflow, entity, or attribute scope, and many-to-many relationships are correct and tested across PostgreSQL and BigQuery.
Standardise the identifier definition across systems, and integration follows automatically. Collision prevention is a structural consequence, not a separate mechanism.
From investing in tech companies to building data platforms - meet the principal data engineer who thinks AI needs unambiguous data models, not a pile of overlapping dbt models.
A practitioner methodology for going from "we need analytics" to a working prototype in hours. Five steps, three groups, one information model. Refined across insurance, automotive, telecom, streaming, and market research.